Feature ranking and rank aggregation for automatic sleep stage classification: a comparative study.

PubMed

Najdi, Shirin; Gharbali, Ali Abdollahi; Fonseca, José Manuel

2017-08-18

Nowadays, sleep quality is one of the most important measures of healthy life, especially considering the huge number of sleep-related disorders. Identifying sleep stages using polysomnographic (PSG) signals is the traditional way of assessing sleep quality. However, the manual process of sleep stage classification is time-consuming, subjective and costly. Therefore, in order to improve the accuracy and efficiency of the sleep stage classification, researchers have been trying to develop automatic classification algorithms. Automatic sleep stage classification mainly consists of three steps: pre-processing, feature extraction and classification. Since classification accuracy is deeply affected by the extracted features, a poor feature vector will adversely affect the classifier and eventually lead to low classification accuracy. Therefore, special attention should be given to the feature extraction and selection process. In this paper the performance of seven feature selection methods, as well as two feature rank aggregation methods, were compared. Pz-Oz EEG, horizontal EOG and submental chin EMG recordings of 22 healthy males and females were used. A comprehensive feature set including 49 features was extracted from these recordings. The extracted features are among the most common and effective features used in sleep stage classification from temporal, spectral, entropy-based and nonlinear categories. The feature selection methods were evaluated and compared using three criteria: classification accuracy, stability, and similarity. Simulation results show that MRMR-MID achieves the highest classification performance while Fisher method provides the most stable ranking. In our simulations, the performance of the aggregation methods was in the average level, although they are known to generate more stable results and better accuracy. The Borda and RRA rank aggregation methods could not outperform significantly the conventional feature ranking methods. Among conventional methods, some of them slightly performed better than others, although the choice of a suitable technique is dependent on the computational complexity and accuracy requirements of the user.

IntelliHealth: A medical decision support application using a novel weighted multi-layer classifier ensemble framework.

PubMed

Bashir, Saba; Qamar, Usman; Khan, Farhan Hassan

2016-02-01

Accuracy plays a vital role in the medical field as it concerns with the life of an individual. Extensive research has been conducted on disease classification and prediction using machine learning techniques. However, there is no agreement on which classifier produces the best results. A specific classifier may be better than others for a specific dataset, but another classifier could perform better for some other dataset. Ensemble of classifiers has been proved to be an effective way to improve classification accuracy. In this research we present an ensemble framework with multi-layer classification using enhanced bagging and optimized weighting. The proposed model called "HM-BagMoov" overcomes the limitations of conventional performance bottlenecks by utilizing an ensemble of seven heterogeneous classifiers. The framework is evaluated on five different heart disease datasets, four breast cancer datasets, two diabetes datasets, two liver disease datasets and one hepatitis dataset obtained from public repositories. The analysis of the results show that ensemble framework achieved the highest accuracy, sensitivity and F-Measure when compared with individual classifiers for all the diseases. In addition to this, the ensemble framework also achieved the highest accuracy when compared with the state of the art techniques. An application named "IntelliHealth" is also developed based on proposed model that may be used by hospitals/doctors for diagnostic advice. Copyright © 2015 Elsevier Inc. All rights reserved.

Optimizing classification performance in an object-based very-high-resolution land use-land cover urban application

NASA Astrophysics Data System (ADS)

Georganos, Stefanos; Grippa, Tais; Vanhuysse, Sabine; Lennert, Moritz; Shimoni, Michal; Wolff, Eléonore

2017-10-01

This study evaluates the impact of three Feature Selection (FS) algorithms in an Object Based Image Analysis (OBIA) framework for Very-High-Resolution (VHR) Land Use-Land Cover (LULC) classification. The three selected FS algorithms, Correlation Based Selection (CFS), Mean Decrease in Accuracy (MDA) and Random Forest (RF) based Recursive Feature Elimination (RFE), were tested on Support Vector Machine (SVM), K-Nearest Neighbor, and Random Forest (RF) classifiers. The results demonstrate that the accuracy of SVM and KNN classifiers are the most sensitive to FS. The RF appeared to be more robust to high dimensionality, although a significant increase in accuracy was found by using the RFE method. In terms of classification accuracy, SVM performed the best using FS, followed by RF and KNN. Finally, only a small number of features is needed to achieve the highest performance using each classifier. This study emphasizes the benefits of rigorous FS for maximizing performance, as well as for minimizing model complexity and interpretation.

Attribute Weighting Based K-Nearest Neighbor Using Gain Ratio

NASA Astrophysics Data System (ADS)

Nababan, A. A.; Sitompul, O. S.; Tulus

2018-04-01

K- Nearest Neighbor (KNN) is a good classifier, but from several studies, the result performance accuracy of KNN still lower than other methods. One of the causes of the low accuracy produced, because each attribute has the same effect on the classification process, while some less relevant characteristics lead to miss-classification of the class assignment for new data. In this research, we proposed Attribute Weighting Based K-Nearest Neighbor Using Gain Ratio as a parameter to see the correlation between each attribute in the data and the Gain Ratio also will be used as the basis for weighting each attribute of the dataset. The accuracy of results is compared to the accuracy acquired from the original KNN method using 10-fold Cross-Validation with several datasets from the UCI Machine Learning repository and KEEL-Dataset Repository, such as abalone, glass identification, haberman, hayes-roth and water quality status. Based on the result of the test, the proposed method was able to increase the classification accuracy of KNN, where the highest difference of accuracy obtained hayes-roth dataset is worth 12.73%, and the lowest difference of accuracy obtained in the abalone dataset of 0.07%. The average result of the accuracy of all dataset increases the accuracy by 5.33%.

Tuberculosis disease diagnosis using artificial immune recognition system.

PubMed

Shamshirband, Shahaboddin; Hessam, Somayeh; Javidnia, Hossein; Amiribesheli, Mohsen; Vahdat, Shaghayegh; Petković, Dalibor; Gani, Abdullah; Kiah, Miss Laiha Mat

2014-01-01

There is a high risk of tuberculosis (TB) disease diagnosis among conventional methods. This study is aimed at diagnosing TB using hybrid machine learning approaches. Patient epicrisis reports obtained from the Pasteur Laboratory in the north of Iran were used. All 175 samples have twenty features. The features are classified based on incorporating a fuzzy logic controller and artificial immune recognition system. The features are normalized through a fuzzy rule based on a labeling system. The labeled features are categorized into normal and tuberculosis classes using the Artificial Immune Recognition Algorithm. Overall, the highest classification accuracy reached was for the 0.8 learning rate (α) values. The artificial immune recognition system (AIRS) classification approaches using fuzzy logic also yielded better diagnosis results in terms of detection accuracy compared to other empirical methods. Classification accuracy was 99.14%, sensitivity 87.00%, and specificity 86.12%.

Diagnosis of diabetes diseases using an Artificial Immune Recognition System2 (AIRS2) with fuzzy K-nearest neighbor.

PubMed

Chikh, Mohamed Amine; Saidi, Meryem; Settouti, Nesma

2012-10-01

The use of expert systems and artificial intelligence techniques in disease diagnosis has been increasing gradually. Artificial Immune Recognition System (AIRS) is one of the methods used in medical classification problems. AIRS2 is a more efficient version of the AIRS algorithm. In this paper, we used a modified AIRS2 called MAIRS2 where we replace the K- nearest neighbors algorithm with the fuzzy K-nearest neighbors to improve the diagnostic accuracy of diabetes diseases. The diabetes disease dataset used in our work is retrieved from UCI machine learning repository. The performances of the AIRS2 and MAIRS2 are evaluated regarding classification accuracy, sensitivity and specificity values. The highest classification accuracy obtained when applying the AIRS2 and MAIRS2 using 10-fold cross-validation was, respectively 82.69% and 89.10%.

Comparison of remote sensing image processing techniques to identify tornado damage areas from Landsat TM data

USGS Publications Warehouse

Myint, S.W.; Yuan, M.; Cerveny, R.S.; Giri, C.P.

2008-01-01

Remote sensing techniques have been shown effective for large-scale damage surveys after a hazardous event in both near real-time or post-event analyses. The paper aims to compare accuracy of common imaging processing techniques to detect tornado damage tracks from Landsat TM data. We employed the direct change detection approach using two sets of images acquired before and after the tornado event to produce a principal component composite images and a set of image difference bands. Techniques in the comparison include supervised classification, unsupervised classification, and objectoriented classification approach with a nearest neighbor classifier. Accuracy assessment is based on Kappa coefficient calculated from error matrices which cross tabulate correctly identified cells on the TM image and commission and omission errors in the result. Overall, the Object-oriented Approach exhibits the highest degree of accuracy in tornado damage detection. PCA and Image Differencing methods show comparable outcomes. While selected PCs can improve detection accuracy 5 to 10%, the Object-oriented Approach performs significantly better with 15-20% higher accuracy than the other two techniques. ?? 2008 by MDPI.

Comparison of Remote Sensing Image Processing Techniques to Identify Tornado Damage Areas from Landsat TM Data

PubMed Central

Myint, Soe W.; Yuan, May; Cerveny, Randall S.; Giri, Chandra P.

2008-01-01

Remote sensing techniques have been shown effective for large-scale damage surveys after a hazardous event in both near real-time or post-event analyses. The paper aims to compare accuracy of common imaging processing techniques to detect tornado damage tracks from Landsat TM data. We employed the direct change detection approach using two sets of images acquired before and after the tornado event to produce a principal component composite images and a set of image difference bands. Techniques in the comparison include supervised classification, unsupervised classification, and object-oriented classification approach with a nearest neighbor classifier. Accuracy assessment is based on Kappa coefficient calculated from error matrices which cross tabulate correctly identified cells on the TM image and commission and omission errors in the result. Overall, the Object-oriented Approach exhibits the highest degree of accuracy in tornado damage detection. PCA and Image Differencing methods show comparable outcomes. While selected PCs can improve detection accuracy 5 to 10%, the Object-oriented Approach performs significantly better with 15-20% higher accuracy than the other two techniques. PMID:27879757

Mediterranean Land Use and Land Cover Classification Assessment Using High Spatial Resolution Data

NASA Astrophysics Data System (ADS)

Elhag, Mohamed; Boteva, Silvena

2016-10-01

Landscape fragmentation is noticeably practiced in Mediterranean regions and imposes substantial complications in several satellite image classification methods. To some extent, high spatial resolution data were able to overcome such complications. For better classification performances in Land Use Land Cover (LULC) mapping, the current research adopts different classification methods comparison for LULC mapping using Sentinel-2 satellite as a source of high spatial resolution. Both of pixel-based and an object-based classification algorithms were assessed; the pixel-based approach employs Maximum Likelihood (ML), Artificial Neural Network (ANN) algorithms, Support Vector Machine (SVM), and, the object-based classification uses the Nearest Neighbour (NN) classifier. Stratified Masking Process (SMP) that integrates a ranking process within the classes based on spectral fluctuation of the sum of the training and testing sites was implemented. An analysis of the overall and individual accuracy of the classification results of all four methods reveals that the SVM classifier was the most efficient overall by distinguishing most of the classes with the highest accuracy. NN succeeded to deal with artificial surface classes in general while agriculture area classes, and forest and semi-natural area classes were segregated successfully with SVM. Furthermore, a comparative analysis indicates that the conventional classification method yielded better accuracy results than the SMP method overall with both classifiers used, ML and SVM.

Improved classification accuracy in 1- and 2-dimensional NMR metabolomics data using the variance stabilising generalised logarithm transformation

PubMed Central

Parsons, Helen M; Ludwig, Christian; Günther, Ulrich L; Viant, Mark R

2007-01-01

Background Classifying nuclear magnetic resonance (NMR) spectra is a crucial step in many metabolomics experiments. Since several multivariate classification techniques depend upon the variance of the data, it is important to first minimise any contribution from unwanted technical variance arising from sample preparation and analytical measurements, and thereby maximise any contribution from wanted biological variance between different classes. The generalised logarithm (glog) transform was developed to stabilise the variance in DNA microarray datasets, but has rarely been applied to metabolomics data. In particular, it has not been rigorously evaluated against other scaling techniques used in metabolomics, nor tested on all forms of NMR spectra including 1-dimensional (1D) 1H, projections of 2D 1H, 1H J-resolved (pJRES), and intact 2D J-resolved (JRES). Results Here, the effects of the glog transform are compared against two commonly used variance stabilising techniques, autoscaling and Pareto scaling, as well as unscaled data. The four methods are evaluated in terms of the effects on the variance of NMR metabolomics data and on the classification accuracy following multivariate analysis, the latter achieved using principal component analysis followed by linear discriminant analysis. For two of three datasets analysed, classification accuracies were highest following glog transformation: 100% accuracy for discriminating 1D NMR spectra of hypoxic and normoxic invertebrate muscle, and 100% accuracy for discriminating 2D JRES spectra of fish livers sampled from two rivers. For the third dataset, pJRES spectra of urine from two breeds of dog, the glog transform and autoscaling achieved equal highest accuracies. Additionally we extended the glog algorithm to effectively suppress noise, which proved critical for the analysis of 2D JRES spectra. Conclusion We have demonstrated that the glog and extended glog transforms stabilise the technical variance in NMR metabolomics datasets. This significantly improves the discrimination between sample classes and has resulted in higher classification accuracies compared to unscaled, autoscaled or Pareto scaled data. Additionally we have confirmed the broad applicability of the glog approach using three disparate datasets from different biological samples using 1D NMR spectra, 1D projections of 2D JRES spectra, and intact 2D JRES spectra. PMID:17605789

Support vector machine for breast cancer classification using diffusion-weighted MRI histogram features: Preliminary study.

PubMed

Vidić, Igor; Egnell, Liv; Jerome, Neil P; Teruel, Jose R; Sjøbakk, Torill E; Østlie, Agnes; Fjøsne, Hans E; Bathen, Tone F; Goa, Pål Erik

2018-05-01

Diffusion-weighted MRI (DWI) is currently one of the fastest developing MRI-based techniques in oncology. Histogram properties from model fitting of DWI are useful features for differentiation of lesions, and classification can potentially be improved by machine learning. To evaluate classification of malignant and benign tumors and breast cancer subtypes using support vector machine (SVM). Prospective. Fifty-one patients with benign (n = 23) and malignant (n = 28) breast tumors (26 ER+, whereof six were HER2+). Patients were imaged with DW-MRI (3T) using twice refocused spin-echo echo-planar imaging with echo time / repetition time (TR/TE) = 9000/86 msec, 90 × 90 matrix size, 2 × 2 mm in-plane resolution, 2.5 mm slice thickness, and 13 b-values. Apparent diffusion coefficient (ADC), relative enhanced diffusivity (RED), and the intravoxel incoherent motion (IVIM) parameters diffusivity (D), pseudo-diffusivity (D*), and perfusion fraction (f) were calculated. The histogram properties (median, mean, standard deviation, skewness, kurtosis) were used as features in SVM (10-fold cross-validation) for differentiation of lesions and subtyping. Accuracies of the SVM classifications were calculated to find the combination of features with highest prediction accuracy. Mann-Whitney tests were performed for univariate comparisons. For benign versus malignant tumors, univariate analysis found 11 histogram properties to be significant differentiators. Using SVM, the highest accuracy (0.96) was achieved from a single feature (mean of RED), or from three feature combinations of IVIM or ADC. Combining features from all models gave perfect classification. No single feature predicted HER2 status of ER + tumors (univariate or SVM), although high accuracy (0.90) was achieved with SVM combining several features. Importantly, these features had to include higher-order statistics (kurtosis and skewness), indicating the importance to account for heterogeneity. Our findings suggest that SVM, using features from a combination of diffusion models, improves prediction accuracy for differentiation of benign versus malignant breast tumors, and may further assist in subtyping of breast cancer. 3 Technical Efficacy: Stage 3 J. Magn. Reson. Imaging 2018;47:1205-1216. © 2017 International Society for Magnetic Resonance in Medicine.

Sub-pixel image classification for forest types in East Texas

NASA Astrophysics Data System (ADS)

Westbrook, Joey

Sub-pixel classification is the extraction of information about the proportion of individual materials of interest within a pixel. Landcover classification at the sub-pixel scale provides more discrimination than traditional per-pixel multispectral classifiers for pixels where the material of interest is mixed with other materials. It allows for the un-mixing of pixels to show the proportion of each material of interest. The materials of interest for this study are pine, hardwood, mixed forest and non-forest. The goal of this project was to perform a sub-pixel classification, which allows a pixel to have multiple labels, and compare the result to a traditional supervised classification, which allows a pixel to have only one label. The satellite image used was a Landsat 5 Thematic Mapper (TM) scene of the Stephen F. Austin Experimental Forest in Nacogdoches County, Texas and the four cover type classes are pine, hardwood, mixed forest and non-forest. Once classified, a multi-layer raster datasets was created that comprised four raster layers where each layer showed the percentage of that cover type within the pixel area. Percentage cover type maps were then produced and the accuracy of each was assessed using a fuzzy error matrix for the sub-pixel classifications, and the results were compared to the supervised classification in which a traditional error matrix was used. The overall accuracy of the sub-pixel classification using the aerial photo for both training and reference data had the highest (65% overall) out of the three sub-pixel classifications. This was understandable because the analyst can visually observe the cover types actually on the ground for training data and reference data, whereas using the FIA (Forest Inventory and Analysis) plot data, the analyst must assume that an entire pixel contains the exact percentage of a cover type found in a plot. An increase in accuracy was found after reclassifying each sub-pixel classification from nine classes with 10 percent interval each to five classes with 20 percent interval each. When compared to the supervised classification which has a satisfactory overall accuracy of 90%, none of the sub-pixel classification achieved the same level. However, since traditional per-pixel classifiers assign only one label to pixels throughout the landscape while sub-pixel classifications assign multiple labels to each pixel, the traditional 85% accuracy of acceptance for pixel-based classifications should not apply to sub-pixel classifications. More research is needed in order to define the level of accuracy that is deemed acceptable for sub-pixel classifications.

Ensemble Methods for Classification of Physical Activities from Wrist Accelerometry.

PubMed

Chowdhury, Alok Kumar; Tjondronegoro, Dian; Chandran, Vinod; Trost, Stewart G

2017-09-01

To investigate whether the use of ensemble learning algorithms improve physical activity recognition accuracy compared to the single classifier algorithms, and to compare the classification accuracy achieved by three conventional ensemble machine learning methods (bagging, boosting, random forest) and a custom ensemble model comprising four algorithms commonly used for activity recognition (binary decision tree, k nearest neighbor, support vector machine, and neural network). The study used three independent data sets that included wrist-worn accelerometer data. For each data set, a four-step classification framework consisting of data preprocessing, feature extraction, normalization and feature selection, and classifier training and testing was implemented. For the custom ensemble, decisions from the single classifiers were aggregated using three decision fusion methods: weighted majority vote, naïve Bayes combination, and behavior knowledge space combination. Classifiers were cross-validated using leave-one subject out cross-validation and compared on the basis of average F1 scores. In all three data sets, ensemble learning methods consistently outperformed the individual classifiers. Among the conventional ensemble methods, random forest models provided consistently high activity recognition; however, the custom ensemble model using weighted majority voting demonstrated the highest classification accuracy in two of the three data sets. Combining multiple individual classifiers using conventional or custom ensemble learning methods can improve activity recognition accuracy from wrist-worn accelerometer data.

Comparison on three classification techniques for sex estimation from the bone length of Asian children below 19 years old: an analysis using different group of ages.

PubMed

Darmawan, M F; Yusuf, Suhaila M; Kadir, M R Abdul; Haron, H

2015-02-01

Sex estimation is used in forensic anthropology to assist the identification of individual remains. However, the estimation techniques tend to be unique and applicable only to a certain population. This paper analyzed sex estimation on living individual child below 19 years old using the length of 19 bones of left hand applied for three classification techniques, which were Discriminant Function Analysis (DFA), Support Vector Machine (SVM) and Artificial Neural Network (ANN) multilayer perceptron. These techniques were carried out on X-ray images of the left hand taken from an Asian population data set. All the 19 bones of the left hand were measured using Free Image software, and all the techniques were performed using MATLAB. The group of age "16-19" years old and "7-9" years old were the groups that could be used for sex estimation with as their average of accuracy percentage was above 80%. ANN model was the best classification technique with the highest average of accuracy percentage in the two groups of age compared to other classification techniques. The results show that each classification technique has the best accuracy percentage on each different group of age. Copyright © 2014 Elsevier Ireland Ltd. All rights reserved.

Multiband tangent space mapping and feature selection for classification of EEG during motor imagery.

PubMed

Islam, Md Rabiul; Tanaka, Toshihisa; Molla, Md Khademul Islam

2018-05-08

When designing multiclass motor imagery-based brain-computer interface (MI-BCI), a so-called tangent space mapping (TSM) method utilizing the geometric structure of covariance matrices is an effective technique. This paper aims to introduce a method using TSM for finding accurate operational frequency bands related brain activities associated with MI tasks. A multichannel electroencephalogram (EEG) signal is decomposed into multiple subbands, and tangent features are then estimated on each subband. A mutual information analysis-based effective algorithm is implemented to select subbands containing features capable of improving motor imagery classification accuracy. Thus obtained features of selected subbands are combined to get feature space. A principal component analysis-based approach is employed to reduce the features dimension and then the classification is accomplished by a support vector machine (SVM). Offline analysis demonstrates the proposed multiband tangent space mapping with subband selection (MTSMS) approach outperforms state-of-the-art methods. It acheives the highest average classification accuracy for all datasets (BCI competition dataset 2a, IIIa, IIIb, and dataset JK-HH1). The increased classification accuracy of MI tasks with the proposed MTSMS approach can yield effective implementation of BCI. The mutual information-based subband selection method is implemented to tune operation frequency bands to represent actual motor imagery tasks.

Mapping of taiga forest units using AIRSAR data and/or optical data, and retrieval of forest parameters

NASA Technical Reports Server (NTRS)

Rignot, Eric; Williams, Cynthia; Way, Jobea; Viereck, Leslie

1993-01-01

A maximum a posteriori Bayesian classifier for multifrequency polarimetric SAR data is used to perform a supervised classification of forest types in the floodplains of Alaska. The image classes include white spruce, balsam poplar, black spruce, alder, non-forests, and open water. The authors investigate the effect on classification accuracy of changing environmental conditions, and of frequency and polarization of the signal. The highest classification accuracy (86 percent correctly classified forest pixels, and 91 percent overall) is obtained combining L- and C-band frequencies fully polarimetric on a date where the forest is just recovering from flooding. The forest map compares favorably with a vegetation map assembled from digitized aerial photos which took five years for completion, and address the state of the forest in 1978, ignoring subsequent fires, changes in the course of the river, clear-cutting of trees, and tree growth. HV-polarization is the most useful polarization at L- and C-band for classification. C-band VV (ERS-1 mode) and L-band HH (J-ERS-1 mode) alone or combined yield unsatisfactory classification accuracies. Additional data acquired in the winter season during thawed and frozen days yield classification accuracies respectively 20 percent and 30 percent lower due to a greater confusion between conifers and deciduous trees. Data acquired at the peak of flooding in May 1991 also yield classification accuracies 10 percent lower because of dominant trunk-ground interactions which mask out finer differences in radar backscatter between tree species. Combination of several of these dates does not improve classification accuracy. For comparison, panchromatic optical data acquired by SPOT in the summer season of 1991 are used to classify the same area. The classification accuracy (78 percent for the forest types and 90 percent if open water is included) is lower than that obtained with AIRSAR although conifers and deciduous trees are better separated due to the presence of leaves on the deciduous trees. Optical data do not separate black spruce and white spruce as well as SAR data, cannot separate alder from balsam poplar, and are of course limited by the frequent cloud cover in the polar regions. Yet, combining SPOT and AIRSAR offers better chances to identify vegetation types independent of ground truth information using a combination of NDVI indexes from SPOT, biomass numbers from AIRSAR, and a segmentation map from either one.

Classification of Parkinson's disease utilizing multi-edit nearest-neighbor and ensemble learning algorithms with speech samples.

PubMed

Zhang, He-Hua; Yang, Liuyang; Liu, Yuchuan; Wang, Pin; Yin, Jun; Li, Yongming; Qiu, Mingguo; Zhu, Xueru; Yan, Fang

2016-11-16

The use of speech based data in the classification of Parkinson disease (PD) has been shown to provide an effect, non-invasive mode of classification in recent years. Thus, there has been an increased interest in speech pattern analysis methods applicable to Parkinsonism for building predictive tele-diagnosis and tele-monitoring models. One of the obstacles in optimizing classifications is to reduce noise within the collected speech samples, thus ensuring better classification accuracy and stability. While the currently used methods are effect, the ability to invoke instance selection has been seldomly examined. In this study, a PD classification algorithm was proposed and examined that combines a multi-edit-nearest-neighbor (MENN) algorithm and an ensemble learning algorithm. First, the MENN algorithm is applied for selecting optimal training speech samples iteratively, thereby obtaining samples with high separability. Next, an ensemble learning algorithm, random forest (RF) or decorrelated neural network ensembles (DNNE), is used to generate trained samples from the collected training samples. Lastly, the trained ensemble learning algorithms are applied to the test samples for PD classification. This proposed method was examined using a more recently deposited public datasets and compared against other currently used algorithms for validation. Experimental results showed that the proposed algorithm obtained the highest degree of improved classification accuracy (29.44%) compared with the other algorithm that was examined. Furthermore, the MENN algorithm alone was found to improve classification accuracy by as much as 45.72%. Moreover, the proposed algorithm was found to exhibit a higher stability, particularly when combining the MENN and RF algorithms. This study showed that the proposed method could improve PD classification when using speech data and can be applied to future studies seeking to improve PD classification methods.

Identification of Anisomerous Motor Imagery EEG Signals Based on Complex Algorithms

PubMed Central

Zhang, Zhiwen; Duan, Feng; Zhou, Xin; Meng, Zixuan

2017-01-01

Motor imagery (MI) electroencephalograph (EEG) signals are widely applied in brain-computer interface (BCI). However, classified MI states are limited, and their classification accuracy rates are low because of the characteristics of nonlinearity and nonstationarity. This study proposes a novel MI pattern recognition system that is based on complex algorithms for classifying MI EEG signals. In electrooculogram (EOG) artifact preprocessing, band-pass filtering is performed to obtain the frequency band of MI-related signals, and then, canonical correlation analysis (CCA) combined with wavelet threshold denoising (WTD) is used for EOG artifact preprocessing. We propose a regularized common spatial pattern (R-CSP) algorithm for EEG feature extraction by incorporating the principle of generic learning. A new classifier combining the K-nearest neighbor (KNN) and support vector machine (SVM) approaches is used to classify four anisomerous states, namely, imaginary movements with the left hand, right foot, and right shoulder and the resting state. The highest classification accuracy rate is 92.5%, and the average classification accuracy rate is 87%. The proposed complex algorithm identification method can significantly improve the identification rate of the minority samples and the overall classification performance. PMID:28874909

Development of a deep convolutional neural network to predict grading of canine meningiomas from magnetic resonance images.

PubMed

Banzato, T; Cherubini, G B; Atzori, M; Zotti, A

2018-05-01

An established deep neural network (DNN) based on transfer learning and a newly designed DNN were tested to predict the grade of meningiomas from magnetic resonance (MR) images in dogs and to determine the accuracy of classification of using pre- and post-contrast T1-weighted (T1W), and T2-weighted (T2W) MR images. The images were randomly assigned to a training set, a validation set and a test set, comprising 60%, 10% and 30% of images, respectively. The combination of DNN and MR sequence displaying the highest discriminating accuracy was used to develop an image classifier to predict the grading of new cases. The algorithm based on transfer learning using the established DNN did not provide satisfactory results, whereas the newly designed DNN had high classification accuracy. On the basis of classification accuracy, an image classifier built on the newly designed DNN using post-contrast T1W images was developed. This image classifier correctly predicted the grading of 8 out of 10 images not included in the data set. Copyright © 2018 The Authors. Published by Elsevier Ltd.. All rights reserved.

Estimation of different data compositions for early-season crop type classification.

PubMed

Hao, Pengyu; Wu, Mingquan; Niu, Zheng; Wang, Li; Zhan, Yulin

2018-01-01

Timely and accurate crop type distribution maps are an important inputs for crop yield estimation and production forecasting as multi-temporal images can observe phenological differences among crops. Therefore, time series remote sensing data are essential for crop type mapping, and image composition has commonly been used to improve the quality of the image time series. However, the optimal composition period is unclear as long composition periods (such as compositions lasting half a year) are less informative and short composition periods lead to information redundancy and missing pixels. In this study, we initially acquired daily 30 m Normalized Difference Vegetation Index (NDVI) time series by fusing MODIS, Landsat, Gaofen and Huanjing (HJ) NDVI, and then composited the NDVI time series using four strategies (daily, 8-day, 16-day, and 32-day). We used Random Forest to identify crop types and evaluated the classification performances of the NDVI time series generated from four composition strategies in two studies regions from Xinjiang, China. Results indicated that crop classification performance improved as crop separabilities and classification accuracies increased, and classification uncertainties dropped in the green-up stage of the crops. When using daily NDVI time series, overall accuracies saturated at 113-day and 116-day in Bole and Luntai, and the saturated overall accuracies (OAs) were 86.13% and 91.89%, respectively. Cotton could be identified 40∼60 days and 35∼45 days earlier than the harvest in Bole and Luntai when using daily, 8-day and 16-day composition NDVI time series since both producer's accuracies (PAs) and user's accuracies (UAs) were higher than 85%. Among the four compositions, the daily NDVI time series generated the highest classification accuracies. Although the 8-day, 16-day and 32-day compositions had similar saturated overall accuracies (around 85% in Bole and 83% in Luntai), the 8-day and 16-day compositions achieved these accuracies around 155-day in Bole and 133-day in Luntai, which were earlier than the 32-day composition (170-day in both Bole and Luntai). Therefore, when the daily NDVI time series cannot be acquired, the 16-day composition is recommended in this study.

Estimation of different data compositions for early-season crop type classification

PubMed Central

Wu, Mingquan; Wang, Li; Zhan, Yulin

2018-01-01

Timely and accurate crop type distribution maps are an important inputs for crop yield estimation and production forecasting as multi-temporal images can observe phenological differences among crops. Therefore, time series remote sensing data are essential for crop type mapping, and image composition has commonly been used to improve the quality of the image time series. However, the optimal composition period is unclear as long composition periods (such as compositions lasting half a year) are less informative and short composition periods lead to information redundancy and missing pixels. In this study, we initially acquired daily 30 m Normalized Difference Vegetation Index (NDVI) time series by fusing MODIS, Landsat, Gaofen and Huanjing (HJ) NDVI, and then composited the NDVI time series using four strategies (daily, 8-day, 16-day, and 32-day). We used Random Forest to identify crop types and evaluated the classification performances of the NDVI time series generated from four composition strategies in two studies regions from Xinjiang, China. Results indicated that crop classification performance improved as crop separabilities and classification accuracies increased, and classification uncertainties dropped in the green-up stage of the crops. When using daily NDVI time series, overall accuracies saturated at 113-day and 116-day in Bole and Luntai, and the saturated overall accuracies (OAs) were 86.13% and 91.89%, respectively. Cotton could be identified 40∼60 days and 35∼45 days earlier than the harvest in Bole and Luntai when using daily, 8-day and 16-day composition NDVI time series since both producer’s accuracies (PAs) and user’s accuracies (UAs) were higher than 85%. Among the four compositions, the daily NDVI time series generated the highest classification accuracies. Although the 8-day, 16-day and 32-day compositions had similar saturated overall accuracies (around 85% in Bole and 83% in Luntai), the 8-day and 16-day compositions achieved these accuracies around 155-day in Bole and 133-day in Luntai, which were earlier than the 32-day composition (170-day in both Bole and Luntai). Therefore, when the daily NDVI time series cannot be acquired, the 16-day composition is recommended in this study. PMID:29868265

AVHRR channel selection for land cover classification

USGS Publications Warehouse

Maxwell, S.K.; Hoffer, R.M.; Chapman, P.L.

2002-01-01

Mapping land cover of large regions often requires processing of satellite images collected from several time periods at many spectral wavelength channels. However, manipulating and processing large amounts of image data increases the complexity and time, and hence the cost, that it takes to produce a land cover map. Very few studies have evaluated the importance of individual Advanced Very High Resolution Radiometer (AVHRR) channels for discriminating cover types, especially the thermal channels (channels 3, 4 and 5). Studies rarely perform a multi-year analysis to determine the impact of inter-annual variability on the classification results. We evaluated 5 years of AVHRR data using combinations of the original AVHRR spectral channels (1-5) to determine which channels are most important for cover type discrimination, yet stabilize inter-annual variability. Particular attention was placed on the channels in the thermal portion of the spectrum. Fourteen cover types over the entire state of Colorado were evaluated using a supervised classification approach on all two-, three-, four- and five-channel combinations for seven AVHRR biweekly composite datasets covering the entire growing season for each of 5 years. Results show that all three of the major portions of the electromagnetic spectrum represented by the AVHRR sensor are required to discriminate cover types effectively and stabilize inter-annual variability. Of the two-channel combinations, channels 1 (red visible) and 2 (near-infrared) had, by far, the highest average overall accuracy (72.2%), yet the inter-annual classification accuracies were highly variable. Including a thermal channel (channel 4) significantly increased the average overall classification accuracy by 5.5% and stabilized interannual variability. Each of the thermal channels gave similar classification accuracies; however, because of the problems in consistently interpreting channel 3 data, either channel 4 or 5 was found to be a more appropriate choice. Substituting the thermal channel with a single elevation layer resulted in equivalent classification accuracies and inter-annual variability.

Sensitivity and specificity of univariate MRI analysis of experimentally degraded cartilage under clinical imaging conditions.

PubMed

Lukas, Vanessa A; Fishbein, Kenneth W; Reiter, David A; Lin, Ping-Chang; Schneider, Erika; Spencer, Richard G

2015-07-01

To evaluate the sensitivity and specificity of classification of pathomimetically degraded bovine nasal cartilage at 3 Tesla and 37°C using univariate MRI measurements of both pure parameter values and intensities of parameter-weighted images. Pre- and posttrypsin degradation values of T1 , T2 , T2 *, magnetization transfer ratio (MTR), and apparent diffusion coefficient (ADC), and corresponding weighted images, were analyzed. Classification based on the Euclidean distance was performed and the quality of classification was assessed through sensitivity, specificity and accuracy (ACC). The classifiers with the highest accuracy values were ADC (ACC = 0.82 ± 0.06), MTR (ACC = 0.78 ± 0.06), T1 (ACC = 0.99 ± 0.01), T2 derived from a three-dimensional (3D) spin-echo sequence (ACC = 0.74 ± 0.05), and T2 derived from a 2D spin-echo sequence (ACC = 0.77 ± 0.06), along with two of the diffusion-weighted signal intensities (b = 333 s/mm(2) : ACC = 0.80 ± 0.05; b = 666 s/mm(2) : ACC = 0.85 ± 0.04). In particular, T1 values differed substantially between the groups, resulting in atypically high classification accuracy. The second-best classifier, diffusion weighting with b = 666 s/mm(2) , as well as all other parameters evaluated, exhibited substantial overlap between pre- and postdegradation groups, resulting in decreased accuracies. Classification according to T1 values showed excellent test characteristics (ACC = 0.99), with several other parameters also showing reasonable performance (ACC > 0.70). Of these, diffusion weighting is particularly promising as a potentially practical clinical modality. As in previous work, we again find that highly statistically significant group mean differences do not necessarily translate into accurate clinical classification rules. © 2014 Wiley Periodicals, Inc.

Automatic Identification of Messages Related to Adverse Drug Reactions from Online User Reviews using Feature-based Classification.

PubMed

Liu, Jingfang; Zhang, Pengzhu; Lu, Yingjie

2014-11-01

User-generated medical messages on Internet contain extensive information related to adverse drug reactions (ADRs) and are known as valuable resources for post-marketing drug surveillance. The aim of this study was to find an effective method to identify messages related to ADRs automatically from online user reviews. We conducted experiments on online user reviews using different feature set and different classification technique. Firstly, the messages from three communities, allergy community, schizophrenia community and pain management community, were collected, the 3000 messages were annotated. Secondly, the N-gram-based features set and medical domain-specific features set were generated. Thirdly, three classification techniques, SVM, C4.5 and Naïve Bayes, were used to perform classification tasks separately. Finally, we evaluated the performance of different method using different feature set and different classification technique by comparing the metrics including accuracy and F-measure. In terms of accuracy, the accuracy of SVM classifier was higher than 0.8, the accuracy of C4.5 classifier or Naïve Bayes classifier was lower than 0.8; meanwhile, the combination feature sets including n-gram-based feature set and domain-specific feature set consistently outperformed single feature set. In terms of F-measure, the highest F-measure is 0.895 which was achieved by using combination feature sets and a SVM classifier. In all, we can get the best classification performance by using combination feature sets and SVM classifier. By using combination feature sets and SVM classifier, we can get an effective method to identify messages related to ADRs automatically from online user reviews.

Forest Cover Estimation in Ireland Using Radar Remote Sensing: A Comparative Analysis of Forest Cover Assessment Methodologies.

PubMed

Devaney, John; Barrett, Brian; Barrett, Frank; Redmond, John; O Halloran, John

2015-01-01

Quantification of spatial and temporal changes in forest cover is an essential component of forest monitoring programs. Due to its cloud free capability, Synthetic Aperture Radar (SAR) is an ideal source of information on forest dynamics in countries with near-constant cloud-cover. However, few studies have investigated the use of SAR for forest cover estimation in landscapes with highly sparse and fragmented forest cover. In this study, the potential use of L-band SAR for forest cover estimation in two regions (Longford and Sligo) in Ireland is investigated and compared to forest cover estimates derived from three national (Forestry2010, Prime2, National Forest Inventory), one pan-European (Forest Map 2006) and one global forest cover (Global Forest Change) product. Two machine-learning approaches (Random Forests and Extremely Randomised Trees) are evaluated. Both Random Forests and Extremely Randomised Trees classification accuracies were high (98.1-98.5%), with differences between the two classifiers being minimal (<0.5%). Increasing levels of post classification filtering led to a decrease in estimated forest area and an increase in overall accuracy of SAR-derived forest cover maps. All forest cover products were evaluated using an independent validation dataset. For the Longford region, the highest overall accuracy was recorded with the Forestry2010 dataset (97.42%) whereas in Sligo, highest overall accuracy was obtained for the Prime2 dataset (97.43%), although accuracies of SAR-derived forest maps were comparable. Our findings indicate that spaceborne radar could aid inventories in regions with low levels of forest cover in fragmented landscapes. The reduced accuracies observed for the global and pan-continental forest cover maps in comparison to national and SAR-derived forest maps indicate that caution should be exercised when applying these datasets for national reporting.

Forest Cover Estimation in Ireland Using Radar Remote Sensing: A Comparative Analysis of Forest Cover Assessment Methodologies

PubMed Central

Devaney, John; Barrett, Brian; Barrett, Frank; Redmond, John; O`Halloran, John

2015-01-01

Quantification of spatial and temporal changes in forest cover is an essential component of forest monitoring programs. Due to its cloud free capability, Synthetic Aperture Radar (SAR) is an ideal source of information on forest dynamics in countries with near-constant cloud-cover. However, few studies have investigated the use of SAR for forest cover estimation in landscapes with highly sparse and fragmented forest cover. In this study, the potential use of L-band SAR for forest cover estimation in two regions (Longford and Sligo) in Ireland is investigated and compared to forest cover estimates derived from three national (Forestry2010, Prime2, National Forest Inventory), one pan-European (Forest Map 2006) and one global forest cover (Global Forest Change) product. Two machine-learning approaches (Random Forests and Extremely Randomised Trees) are evaluated. Both Random Forests and Extremely Randomised Trees classification accuracies were high (98.1–98.5%), with differences between the two classifiers being minimal (<0.5%). Increasing levels of post classification filtering led to a decrease in estimated forest area and an increase in overall accuracy of SAR-derived forest cover maps. All forest cover products were evaluated using an independent validation dataset. For the Longford region, the highest overall accuracy was recorded with the Forestry2010 dataset (97.42%) whereas in Sligo, highest overall accuracy was obtained for the Prime2 dataset (97.43%), although accuracies of SAR-derived forest maps were comparable. Our findings indicate that spaceborne radar could aid inventories in regions with low levels of forest cover in fragmented landscapes. The reduced accuracies observed for the global and pan-continental forest cover maps in comparison to national and SAR-derived forest maps indicate that caution should be exercised when applying these datasets for national reporting. PMID:26262681

Classification of polycystic ovary based on ultrasound images using competitive neural network

NASA Astrophysics Data System (ADS)

Dewi, R. M.; Adiwijaya; Wisesty, U. N.; Jondri

2018-03-01

Infertility in the women reproduction system due to inhibition of follicles maturation process causing the number of follicles which is called polycystic ovaries (PCO). PCO detection is still operated manually by a gynecologist by counting the number and size of follicles in the ovaries, so it takes a long time and needs high accuracy. In general, PCO can be detected by calculating stereology or feature extraction and classification. In this paper, we designed a system to classify PCO by using the feature extraction (Gabor Wavelet method) and Competitive Neural Network (CNN). CNN was selected because this method is the combination between Hemming Net and The Max Net so that the data classification can be performed based on the specific characteristics of ultrasound data. Based on the result of system testing, Competitive Neural Network obtained the highest accuracy is 80.84% and the time process is 60.64 seconds (when using 32 feature vectors as well as weight and bias values respectively of 0.03 and 0.002).

A classification of marked hijaiyah letters' pronunciation using hidden Markov model

NASA Astrophysics Data System (ADS)

Wisesty, Untari N.; Mubarok, M. Syahrul; Adiwijaya

2017-08-01

Hijaiyah letters are the letters that arrange the words in Al Qur'an consisting of 28 letters. They symbolize the consonant sounds. On the other hand, the vowel sounds are symbolized by harokat/marks. Speech recognition system is a system used to process the sound signal to be data so that it can be recognized by computer. To build the system, some stages are needed i.e characteristics/feature extraction and classification. In this research, LPC and MFCC extraction method, K-Means Quantization vector and Hidden Markov Model classification are used. The data used are the 28 letters and 6 harakat with the total class of 168. After several are testing done, it can be concluded that the system can recognize the pronunciation pattern of marked hijaiyah letter very well in the training data with its highest accuracy of 96.1% using the feature of LPC extraction and 94% using the MFCC. Meanwhile, when testing system is used, the accuracy decreases up to 41%.

Low-back electromyography (EMG) data-driven load classification for dynamic lifting tasks.

PubMed

Totah, Deema; Ojeda, Lauro; Johnson, Daniel D; Gates, Deanna; Mower Provost, Emily; Barton, Kira

2018-01-01

Numerous devices have been designed to support the back during lifting tasks. To improve the utility of such devices, this research explores the use of preparatory muscle activity to classify muscle loading and initiate appropriate device activation. The goal of this study was to determine the earliest time window that enabled accurate load classification during a dynamic lifting task. Nine subjects performed thirty symmetrical lifts, split evenly across three weight conditions (no-weight, 10-lbs and 24-lbs), while low-back muscle activity data was collected. Seven descriptive statistics features were extracted from 100 ms windows of data. A multinomial logistic regression (MLR) classifier was trained and tested, employing leave-one subject out cross-validation, to classify lifted load values. Dimensionality reduction was achieved through feature cross-correlation analysis and greedy feedforward selection. The time of full load support by the subject was defined as load-onset. Regions of highest average classification accuracy started at 200 ms before until 200 ms after load-onset with average accuracies ranging from 80% (±10%) to 81% (±7%). The average recall for each class ranged from 69-92%. These inter-subject classification results indicate that preparatory muscle activity can be leveraged to identify the intent to lift a weight up to 100 ms prior to load-onset. The high accuracies shown indicate the potential to utilize intent classification for assistive device applications. Active assistive devices, e.g. exoskeletons, could prevent back injury by off-loading low-back muscles. Early intent classification allows more time for actuators to respond and integrate seamlessly with the user.

Factors That Affect Large Subunit Ribosomal DNA Amplicon Sequencing Studies of Fungal Communities: Classification Method, Primer Choice, and Error

PubMed Central

Porter, Teresita M.; Golding, G. Brian

2012-01-01

Nuclear large subunit ribosomal DNA is widely used in fungal phylogenetics and to an increasing extent also amplicon-based environmental sequencing. The relatively short reads produced by next-generation sequencing, however, makes primer choice and sequence error important variables for obtaining accurate taxonomic classifications. In this simulation study we tested the performance of three classification methods: 1) a similarity-based method (BLAST + Metagenomic Analyzer, MEGAN); 2) a composition-based method (Ribosomal Database Project naïve Bayesian classifier, NBC); and, 3) a phylogeny-based method (Statistical Assignment Package, SAP). We also tested the effects of sequence length, primer choice, and sequence error on classification accuracy and perceived community composition. Using a leave-one-out cross validation approach, results for classifications to the genus rank were as follows: BLAST + MEGAN had the lowest error rate and was particularly robust to sequence error; SAP accuracy was highest when long LSU query sequences were classified; and, NBC runs significantly faster than the other tested methods. All methods performed poorly with the shortest 50–100 bp sequences. Increasing simulated sequence error reduced classification accuracy. Community shifts were detected due to sequence error and primer selection even though there was no change in the underlying community composition. Short read datasets from individual primers, as well as pooled datasets, appear to only approximate the true community composition. We hope this work informs investigators of some of the factors that affect the quality and interpretation of their environmental gene surveys. PMID:22558215

Multi-class biological tissue classification based on a multi-classifier: Preliminary study of an automatic output power control for ultrasonic surgical units.

PubMed

Youn, Su Hyun; Sim, Taeyong; Choi, Ahnryul; Song, Jinsung; Shin, Ki Young; Lee, Il Kwon; Heo, Hyun Mu; Lee, Daeweon; Mun, Joung Hwan

2015-06-01

Ultrasonic surgical units (USUs) have the advantage of minimizing tissue damage during surgeries that require tissue dissection by reducing problems such as coagulation and unwanted carbonization, but the disadvantage of requiring manual adjustment of power output according to the target tissue. In order to overcome this limitation, it is necessary to determine the properties of in vivo tissues automatically. We propose a multi-classifier that can accurately classify tissues based on the unique impedance of each tissue. For this purpose, a multi-classifier was built based on single classifiers with high classification rates, and the classification accuracy of the proposed model was compared with that of single classifiers for various electrode types (Type-I: 6 mm invasive; Type-II: 3 mm invasive; Type-III: surface). The sensitivity and positive predictive value (PPV) of the multi-classifier by cross checks were determined. According to the 10-fold cross validation results, the classification accuracy of the proposed model was significantly higher (p<0.05 or <0.01) than that of existing single classifiers for all electrode types. In particular, the classification accuracy of the proposed model was highest when the 3mm invasive electrode (Type-II) was used (sensitivity=97.33-100.00%; PPV=96.71-100.00%). The results of this study are an important contribution to achieving automatic optimal output power adjustment of USUs according to the properties of individual tissues. Copyright © 2015 Elsevier Ltd. All rights reserved.

Determination of Classification Accuracy for Land Use/cover Types Using Landsat-Tm Spot-Mss and Multipolarized and Multi-Channel Synthetic Aperture Radar

NASA Astrophysics Data System (ADS)

Dondurur, Mehmet

The primary objective of this study was to determine the degree to which modern SAR systems can be used to obtain information about the Earth's vegetative resources. Information obtainable from microwave synthetic aperture radar (SAR) data was compared with that obtainable from LANDSAT-TM and SPOT data. Three hypotheses were tested: (a) Classification of land cover/use from SAR data can be accomplished on a pixel-by-pixel basis with the same overall accuracy as from LANDSAT-TM and SPOT data. (b) Classification accuracy for individual land cover/use classes will differ between sensors. (c) Combining information derived from optical and SAR data into an integrated monitoring system will improve overall and individual land cover/use class accuracies. The study was conducted with three data sets for the Sleeping Bear Dunes test site in the northwestern part of Michigan's lower peninsula, including an October 1982 LANDSAT-TM scene, a June 1989 SPOT scene and C-, L- and P-Band radar data from the Jet Propulsion Laboratory AIRSAR. Reference data were derived from the Michigan Resource Information System (MIRIS) and available color infrared aerial photos. Classification and rectification of data sets were done using ERDAS Image Processing Programs. Classification algorithms included Maximum Likelihood, Mahalanobis Distance, Minimum Spectral Distance, ISODATA, Parallelepiped, and Sequential Cluster Analysis. Classified images were rectified as necessary so that all were at the same scale and oriented north-up. Results were analyzed with contingency tables and percent correctly classified (PCC) and Cohen's Kappa (CK) as accuracy indices using CSLANT and ImagePro programs developed for this study. Accuracy analyses were based upon a 1.4 by 6.5 km area with its long axis east-west. Reference data for this subscene total 55,770 15 by 15 m pixels with sixteen cover types, including seven level III forest classes, three level III urban classes, two level II range classes, two water classes, one wetland class and one agriculture class. An initial analysis was made without correcting the 1978 MIRIS reference data to the different dates of the TM, SPOT and SAR data sets. In this analysis, highest overall classification accuracy (PCC) was 87% with the TM data set, with both SPOT and C-Band SAR at 85%, a difference statistically significant at the 0.05 level. When the reference data were corrected for land cover change between 1978 and 1991, classification accuracy with the C-Band SAR data increased to 87%. Classification accuracy differed from sensor to sensor for individual land cover classes, Combining sensors into hypothetical multi-sensor systems resulted in higher accuracies than for any single sensor. Combining LANDSAT -TM and C-Band SAR yielded an overall classification accuracy (PCC) of 92%. The results of this study indicate that C-Band SAR data provide an acceptable substitute for LANDSAT-TM or SPOT data when land cover information is desired of areas where cloud cover obscures the terrain. Even better results can be obtained by integrating TM and C-Band SAR data into a multi-sensor system.

A machine learning approach to multi-level ECG signal quality classification.

PubMed

Li, Qiao; Rajagopalan, Cadathur; Clifford, Gari D

2014-12-01

Current electrocardiogram (ECG) signal quality assessment studies have aimed to provide a two-level classification: clean or noisy. However, clinical usage demands more specific noise level classification for varying applications. This work outlines a five-level ECG signal quality classification algorithm. A total of 13 signal quality metrics were derived from segments of ECG waveforms, which were labeled by experts. A support vector machine (SVM) was trained to perform the classification and tested on a simulated dataset and was validated using data from the MIT-BIH arrhythmia database (MITDB). The simulated training and test datasets were created by selecting clean segments of the ECG in the 2011 PhysioNet/Computing in Cardiology Challenge database, and adding three types of real ECG noise at different signal-to-noise ratio (SNR) levels from the MIT-BIH Noise Stress Test Database (NSTDB). The MITDB was re-annotated for five levels of signal quality. Different combinations of the 13 metrics were trained and tested on the simulated datasets and the best combination that produced the highest classification accuracy was selected and validated on the MITDB. Performance was assessed using classification accuracy (Ac), and a single class overlap accuracy (OAc), which assumes that an individual type classified into an adjacent class is acceptable. An Ac of 80.26% and an OAc of 98.60% on the test set were obtained by selecting 10 metrics while 57.26% (Ac) and 94.23% (OAc) were the numbers for the unseen MITDB validation data without retraining. By performing the fivefold cross validation, an Ac of 88.07±0.32% and OAc of 99.34±0.07% were gained on the validation fold of MITDB. Copyright © 2014 Elsevier Ireland Ltd. All rights reserved.

Testing a tool for the classification of study designs in systematic reviews of interventions and exposures showed moderate reliability and low accuracy.

PubMed

Hartling, Lisa; Bond, Kenneth; Santaguida, P Lina; Viswanathan, Meera; Dryden, Donna M

2011-08-01

To develop and test a study design classification tool. We contacted relevant organizations and individuals to identify tools used to classify study designs and ranked these using predefined criteria. The highest ranked tool was a design algorithm developed, but no longer advocated, by the Cochrane Non-Randomized Studies Methods Group; this was modified to include additional study designs and decision points. We developed a reference classification for 30 studies; 6 testers applied the tool to these studies. Interrater reliability (Fleiss' κ) and accuracy against the reference classification were assessed. The tool was further revised and retested. Initial reliability was fair among the testers (κ=0.26) and the reference standard raters κ=0.33). Testing after revisions showed improved reliability (κ=0.45, moderate agreement) with improved, but still low, accuracy. The most common disagreements were whether the study design was experimental (5 of 15 studies), and whether there was a comparison of any kind (4 of 15 studies). Agreement was higher among testers who had completed graduate level training versus those who had not. The moderate reliability and low accuracy may be because of lack of clarity and comprehensiveness of the tool, inadequate reporting of the studies, and variability in tester characteristics. The results may not be generalizable to all published studies, as the test studies were selected because they had posed challenges for previous reviewers with respect to their design classification. Application of such a tool should be accompanied by training, pilot testing, and context-specific decision rules. Copyright © 2011 Elsevier Inc. All rights reserved.

An assessment of commonly employed satellite-based remote sensors for mapping mangrove species in Mexico using an NDVI-based classification scheme.

PubMed

Valderrama-Landeros, L; Flores-de-Santiago, F; Kovacs, J M; Flores-Verdugo, F

2017-12-14

Optimizing the classification accuracy of a mangrove forest is of utmost importance for conservation practitioners. Mangrove forest mapping using satellite-based remote sensing techniques is by far the most common method of classification currently used given the logistical difficulties of field endeavors in these forested wetlands. However, there is now an abundance of options from which to choose in regards to satellite sensors, which has led to substantially different estimations of mangrove forest location and extent with particular concern for degraded systems. The objective of this study was to assess the accuracy of mangrove forest classification using different remotely sensed data sources (i.e., Landsat-8, SPOT-5, Sentinel-2, and WorldView-2) for a system located along the Pacific coast of Mexico. Specifically, we examined a stressed semiarid mangrove forest which offers a variety of conditions such as dead areas, degraded stands, healthy mangroves, and very dense mangrove island formations. The results indicated that Landsat-8 (30 m per pixel) had  the lowest overall accuracy at 64% and that WorldView-2 (1.6 m per pixel) had the highest at 93%. Moreover, the SPOT-5 and the Sentinel-2 classifications (10 m per pixel) were very similar having accuracies of 75 and 78%, respectively. In comparison to WorldView-2, the other sensors overestimated the extent of Laguncularia racemosa and underestimated the extent of Rhizophora mangle. When considering such type of sensors, the higher spatial resolution can be particularly important in mapping small mangrove islands that often occur in degraded mangrove systems.

ICA-Based Imagined Conceptual Words Classification on EEG Signals.

PubMed

Imani, Ehsan; Pourmohammad, Ali; Bagheri, Mahsa; Mobasheri, Vida

2017-01-01

Independent component analysis (ICA) has been used for detecting and removing the eye artifacts conventionally. However, in this research, it was used not only for detecting the eye artifacts, but also for detecting the brain-produced signals of two conceptual danger and information category words. In this cross-sectional research, electroencephalography (EEG) signals were recorded using Micromed and 19-channel helmet devices in unipolar mode, wherein Cz electrode was selected as the reference electrode. In the first part of this research, the statistical community test case included four men and four women, who were 25-30 years old. In the designed task, three groups of traffic signs were considered, in which two groups referred to the concept of danger, and the third one referred to the concept of information. In the second part, the three volunteers, two men and one woman, who had the best results, were chosen from among eight participants. In the second designed task, direction arrows (up, down, left, and right) were used. For the 2/8 volunteers in the rest times, very high-power alpha waves were observed from the back of the head; however, in the thinking times, they were different. According to this result, alpha waves for changing the task from thinking to rest condition took at least 3 s for the two volunteers, and it was at most 5 s until they went to the absolute rest condition. For the 7/8 volunteers, the danger and information signals were well classified; these differences for the 5/8 volunteers were observed in the right hemisphere, and, for the other three volunteers, the differences were observed in the left hemisphere. For the second task, simulations showed that the best classification accuracies resulted when the time window was 2.5 s. In addition, it also showed that the features of the autoregressive (AR)-15 model coefficients were the best choices for extracting the features. For all the states of neural network except hardlim discriminator function, the classification accuracies were almost the same and not very different. Linear discriminant analysis (LDA) in comparison with the neural network yielded higher classification accuracies. ICA is a suitable algorithm for recognizing of the word's concept and its place in the brain. Achieved results from this experiment were the same compared with the results from other methods such as functional magnetic resonance imaging and methods based on the brain signals (EEG) in the vowel imagination and covert speech. Herein, the highest classification accuracy was obtained by extracting the target signal from the output of the ICA and extracting the features of coefficients AR model with time interval of 2.5 s. Finally, LDA resulted in the highest classification accuracy more than 60%.

Comparative analysis of Worldview-2 and Landsat 8 for coastal saltmarsh mapping accuracy assessment

NASA Astrophysics Data System (ADS)

Rasel, Sikdar M. M.; Chang, Hsing-Chung; Diti, Israt Jahan; Ralph, Tim; Saintilan, Neil

2016-05-01

Coastal saltmarsh and their constituent components and processes are of an interest scientifically due to their ecological function and services. However, heterogeneity and seasonal dynamic of the coastal wetland system makes it challenging to map saltmarshes with remotely sensed data. This study selected four important saltmarsh species Pragmitis australis, Sporobolus virginicus, Ficiona nodosa and Schoeloplectus sp. as well as a Mangrove and Pine tree species, Avecinia and Casuarina sp respectively. High Spatial Resolution Worldview-2 data and Coarse Spatial resolution Landsat 8 imagery were selected in this study. Among the selected vegetation types some patches ware fragmented and close to the spatial resolution of Worldview-2 data while and some patch were larger than the 30 meter resolution of Landsat 8 data. This study aims to test the effectiveness of different classifier for the imagery with various spatial and spectral resolutions. Three different classification algorithm, Maximum Likelihood Classifier (MLC), Support Vector Machine (SVM) and Artificial Neural Network (ANN) were tested and compared with their mapping accuracy of the results derived from both satellite imagery. For Worldview-2 data SVM was giving the higher overall accuracy (92.12%, kappa =0.90) followed by ANN (90.82%, Kappa 0.89) and MLC (90.55%, kappa = 0.88). For Landsat 8 data, MLC (82.04%) showed the highest classification accuracy comparing to SVM (77.31%) and ANN (75.23%). The producer accuracy of the classification results were also presented in the paper.

Multispectral imaging burn wound tissue classification system: a comparison of test accuracies between several common machine learning algorithms

NASA Astrophysics Data System (ADS)

Squiers, John J.; Li, Weizhi; King, Darlene R.; Mo, Weirong; Zhang, Xu; Lu, Yang; Sellke, Eric W.; Fan, Wensheng; DiMaio, J. Michael; Thatcher, Jeffrey E.

2016-03-01

The clinical judgment of expert burn surgeons is currently the standard on which diagnostic and therapeutic decisionmaking regarding burn injuries is based. Multispectral imaging (MSI) has the potential to increase the accuracy of burn depth assessment and the intraoperative identification of viable wound bed during surgical debridement of burn injuries. A highly accurate classification model must be developed using machine-learning techniques in order to translate MSI data into clinically-relevant information. An animal burn model was developed to build an MSI training database and to study the burn tissue classification ability of several models trained via common machine-learning algorithms. The algorithms tested, from least to most complex, were: K-nearest neighbors (KNN), decision tree (DT), linear discriminant analysis (LDA), weighted linear discriminant analysis (W-LDA), quadratic discriminant analysis (QDA), ensemble linear discriminant analysis (EN-LDA), ensemble K-nearest neighbors (EN-KNN), and ensemble decision tree (EN-DT). After the ground-truth database of six tissue types (healthy skin, wound bed, blood, hyperemia, partial injury, full injury) was generated by histopathological analysis, we used 10-fold cross validation to compare the algorithms' performances based on their accuracies in classifying data against the ground truth, and each algorithm was tested 100 times. The mean test accuracy of the algorithms were KNN 68.3%, DT 61.5%, LDA 70.5%, W-LDA 68.1%, QDA 68.9%, EN-LDA 56.8%, EN-KNN 49.7%, and EN-DT 36.5%. LDA had the highest test accuracy, reflecting the bias-variance tradeoff over the range of complexities inherent to the algorithms tested. Several algorithms were able to match the current standard in burn tissue classification, the clinical judgment of expert burn surgeons. These results will guide further development of an MSI burn tissue classification system. Given that there are few surgeons and facilities specializing in burn care, this technology may improve the standard of burn care for patients without access to specialized facilities.

Accurate, Rapid Taxonomic Classification of Fungal Large-Subunit rRNA Genes

PubMed Central

Liu, Kuan-Liang; Porras-Alfaro, Andrea; Eichorst, Stephanie A.

2012-01-01

Taxonomic and phylogenetic fingerprinting based on sequence analysis of gene fragments from the large-subunit rRNA (LSU) gene or the internal transcribed spacer (ITS) region is becoming an integral part of fungal classification. The lack of an accurate and robust classification tool trained by a validated sequence database for taxonomic placement of fungal LSU genes is a severe limitation in taxonomic analysis of fungal isolates or large data sets obtained from environmental surveys. Using a hand-curated set of 8,506 fungal LSU gene fragments, we determined the performance characteristics of a naïve Bayesian classifier across multiple taxonomic levels and compared the classifier performance to that of a sequence similarity-based (BLASTN) approach. The naïve Bayesian classifier was computationally more rapid (>460-fold with our system) than the BLASTN approach, and it provided equal or superior classification accuracy. Classifier accuracies were compared using sequence fragments of 100 bp and 400 bp and two different PCR primer anchor points to mimic sequence read lengths commonly obtained using current high-throughput sequencing technologies. Accuracy was higher with 400-bp sequence reads than with 100-bp reads. It was also significantly affected by sequence location across the 1,400-bp test region. The highest accuracy was obtained across either the D1 or D2 variable region. The naïve Bayesian classifier provides an effective and rapid means to classify fungal LSU sequences from large environmental surveys. The training set and tool are publicly available through the Ribosomal Database Project (http://rdp.cme.msu.edu/classifier/classifier.jsp). PMID:22194300

Estimating workload using EEG spectral power and ERPs in the n-back task

NASA Astrophysics Data System (ADS)

Brouwer, Anne-Marie; Hogervorst, Maarten A.; van Erp, Jan B. F.; Heffelaar, Tobias; Zimmerman, Patrick H.; Oostenveld, Robert

2012-08-01

Previous studies indicate that both electroencephalogram (EEG) spectral power (in particular the alpha and theta band) and event-related potentials (ERPs) (in particular the P300) can be used as a measure of mental work or memory load. We compare their ability to estimate workload level in a well-controlled task. In addition, we combine both types of measures in a single classification model to examine whether this results in higher classification accuracy than either one alone. Participants watched a sequence of visually presented letters and indicated whether or not the current letter was the same as the one (n instances) before. Workload was varied by varying n. We developed different classification models using ERP features, frequency power features or a combination (fusion). Training and testing of the models simulated an online workload estimation situation. All our ERP, power and fusion models provide classification accuracies between 80% and 90% when distinguishing between the highest and the lowest workload condition after 2 min. For 32 out of 35 participants, classification was significantly higher than chance level after 2.5 s (or one letter) as estimated by the fusion model. Differences between the models are rather small, though the fusion model performs better than the other models when only short data segments are available for estimating workload.

Bloodstain pattern classification: Accuracy, effect of contextual information and the role of analyst characteristics.

PubMed

Osborne, Nikola K P; Taylor, Michael C; Healey, Matthew; Zajac, Rachel

2016-03-01

It is becoming increasingly apparent that contextual information can exert a considerable influence on decisions about forensic evidence. Here, we explored accuracy and contextual influence in bloodstain pattern classification, and how these variables might relate to analyst characteristics. Thirty-nine bloodstain pattern analysts with varying degrees of experience each completed measures of compliance, decision-making style, and need for closure. Analysts then examined a bloodstain pattern without any additional contextual information, and allocated votes to listed pattern types according to favoured and less favoured classifications. Next, if they believed it would assist with their classification, analysts could request items of contextual information - from commonly encountered sources of information in bloodstain pattern analysis - and update their vote allocation. We calculated a shift score for each item of contextual information based on vote reallocation. Almost all forms of contextual information influenced decision-making, with medical findings leading to the highest shift scores. Although there was a small positive association between shift scores and the degree to which analysts displayed an intuitive decision-making style, shift scores did not vary meaningfully as a function of experience or the other characteristics measured. Almost all of the erroneous classifications were made by novice analysts. Copyright © 2016 The Chartered Society of Forensic Sciences. Published by Elsevier Ireland Ltd. All rights reserved.

Low-back electromyography (EMG) data-driven load classification for dynamic lifting tasks

PubMed Central

Ojeda, Lauro; Johnson, Daniel D.; Gates, Deanna; Mower Provost, Emily; Barton, Kira

2018-01-01

Objective Numerous devices have been designed to support the back during lifting tasks. To improve the utility of such devices, this research explores the use of preparatory muscle activity to classify muscle loading and initiate appropriate device activation. The goal of this study was to determine the earliest time window that enabled accurate load classification during a dynamic lifting task. Methods Nine subjects performed thirty symmetrical lifts, split evenly across three weight conditions (no-weight, 10-lbs and 24-lbs), while low-back muscle activity data was collected. Seven descriptive statistics features were extracted from 100 ms windows of data. A multinomial logistic regression (MLR) classifier was trained and tested, employing leave-one subject out cross-validation, to classify lifted load values. Dimensionality reduction was achieved through feature cross-correlation analysis and greedy feedforward selection. The time of full load support by the subject was defined as load-onset. Results Regions of highest average classification accuracy started at 200 ms before until 200 ms after load-onset with average accuracies ranging from 80% (±10%) to 81% (±7%). The average recall for each class ranged from 69–92%. Conclusion These inter-subject classification results indicate that preparatory muscle activity can be leveraged to identify the intent to lift a weight up to 100 ms prior to load-onset. The high accuracies shown indicate the potential to utilize intent classification for assistive device applications. Significance Active assistive devices, e.g. exoskeletons, could prevent back injury by off-loading low-back muscles. Early intent classification allows more time for actuators to respond and integrate seamlessly with the user. PMID:29447252

Classification of Focal and Non Focal Epileptic Seizures Using Multi-Features and SVM Classifier.

PubMed

Sriraam, N; Raghu, S

2017-09-02

Identifying epileptogenic zones prior to surgery is an essential and crucial step in treating patients having pharmacoresistant focal epilepsy. Electroencephalogram (EEG) is a significant measurement benchmark to assess patients suffering from epilepsy. This paper investigates the application of multi-features derived from different domains to recognize the focal and non focal epileptic seizures obtained from pharmacoresistant focal epilepsy patients from Bern Barcelona database. From the dataset, five different classification tasks were formed. Total 26 features were extracted from focal and non focal EEG. Significant features were selected using Wilcoxon rank sum test by setting p-value (p < 0.05) and z-score (-1.96 > z > 1.96) at 95% significance interval. Hypothesis was made that the effect of removing outliers improves the classification accuracy. Turkey's range test was adopted for pruning outliers from feature set. Finally, 21 features were classified using optimized support vector machine (SVM) classifier with 10-fold cross validation. Bayesian optimization technique was adopted to minimize the cross-validation loss. From the simulation results, it was inferred that the highest sensitivity, specificity, and classification accuracy of 94.56%, 89.74%, and 92.15% achieved respectively and found to be better than the state-of-the-art approaches. Further, it was observed that the classification accuracy improved from 80.2% with outliers to 92.15% without outliers. The classifier performance metrics ensures the suitability of the proposed multi-features with optimized SVM classifier. It can be concluded that the proposed approach can be applied for recognition of focal EEG signals to localize epileptogenic zones.

Landscape object-based analysis of wetland plant functional types: the effects of spatial scale, vegetation classes and classifier methods

NASA Astrophysics Data System (ADS)

Dronova, I.; Gong, P.; Wang, L.; Clinton, N.; Fu, W.; Qi, S.

2011-12-01

Remote sensing-based vegetation classifications representing plant function such as photosynthesis and productivity are challenging in wetlands with complex cover and difficult field access. Recent advances in object-based image analysis (OBIA) and machine-learning algorithms offer new classification tools; however, few comparisons of different algorithms and spatial scales have been discussed to date. We applied OBIA to delineate wetland plant functional types (PFTs) for Poyang Lake, the largest freshwater lake in China and Ramsar wetland conservation site, from 30-m Landsat TM scene at the peak of spring growing season. We targeted major PFTs (C3 grasses, C3 forbs and different types of C4 grasses and aquatic vegetation) that are both key players in system's biogeochemical cycles and critical providers of waterbird habitat. Classification results were compared among: a) several object segmentation scales (with average object sizes 900-9000 m2); b) several families of statistical classifiers (including Bayesian, Logistic, Neural Network, Decision Trees and Support Vector Machines) and c) two hierarchical levels of vegetation classification, a generalized 3-class set and more detailed 6-class set. We found that classification benefited from object-based approach which allowed including object shape, texture and context descriptors in classification. While a number of classifiers achieved high accuracy at the finest pixel-equivalent segmentation scale, the highest accuracies and best agreement among algorithms occurred at coarser object scales. No single classifier was consistently superior across all scales, although selected algorithms of Neural Network, Logistic and K-Nearest Neighbors families frequently provided the best discrimination of classes at different scales. The choice of vegetation categories also affected classification accuracy. The 6-class set allowed for higher individual class accuracies but lower overall accuracies than the 3-class set because individual classes differed in scales at which they were best discriminated from others. Main classification challenges included a) presence of C3 grasses in C4-grass areas, particularly following harvesting of C4 reeds and b) mixtures of emergent, floating and submerged aquatic plants at sub-object and sub-pixel scales. We conclude that OBIA with advanced statistical classifiers offers useful instruments for landscape vegetation analyses, and that spatial scale considerations are critical in mapping PFTs, while multi-scale comparisons can be used to guide class selection. Future work will further apply fuzzy classification and field-collected spectral data for PFT analysis and compare results with MODIS PFT products.

Towards automatic lithological classification from remote sensing data using support vector machines

NASA Astrophysics Data System (ADS)

Yu, Le; Porwal, Alok; Holden, Eun-Jung; Dentith, Michael

2010-05-01

Remote sensing data can be effectively used as a mean to build geological knowledge for poorly mapped terrains. Spectral remote sensing data from space- and air-borne sensors have been widely used to geological mapping, especially in areas of high outcrop density in arid regions. However, spectral remote sensing information by itself cannot be efficiently used for a comprehensive lithological classification of an area due to (1) diagnostic spectral response of a rock within an image pixel is conditioned by several factors including the atmospheric effects, spectral and spatial resolution of the image, sub-pixel level heterogeneity in chemical and mineralogical composition of the rock, presence of soil and vegetation cover; (2) only surface information and is therefore highly sensitive to the noise due to weathering, soil cover, and vegetation. Consequently, for efficient lithological classification, spectral remote sensing data needs to be supplemented with other remote sensing datasets that provide geomorphological and subsurface geological information, such as digital topographic model (DEM) and aeromagnetic data. Each of the datasets contain significant information about geology that, in conjunction, can potentially be used for automated lithological classification using supervised machine learning algorithms. In this study, support vector machine (SVM), which is a kernel-based supervised learning method, was applied to automated lithological classification of a study area in northwestern India using remote sensing data, namely, ASTER, DEM and aeromagnetic data. Several digital image processing techniques were used to produce derivative datasets that contained enhanced information relevant to lithological discrimination. A series of SVMs (trained using k-folder cross-validation with grid search) were tested using various combinations of input datasets selected from among 50 datasets including the original 14 ASTER bands and 36 derivative datasets (including 14 principal component bands, 14 independent component bands, 3 band ratios, 3 DEM derivatives: slope/curvatureroughness and 2 aeromagnetic derivatives: mean and variance of susceptibility) extracted from the ASTER, DEM and aeromagnetic data, in order to determine the optimal inputs that provide the highest classification accuracy. It was found that a combination of ASTER-derived independent components, principal components and band ratios, DEM-derived slope, curvature and roughness, and aeromagnetic-derived mean and variance of magnetic susceptibility provide the highest classification accuracy of 93.4% on independent test samples. A comparison of the classification results of the SVM with those of maximum likelihood (84.9%) and minimum distance (38.4%) classifiers clearly show that the SVM algorithm returns much higher classification accuracy. Therefore, the SVM method can be used to produce quick and reliable geological maps from scarce geological information, which is still the case with many under-developed frontier regions of the world.

Structural MRI in Frontotemporal Dementia: Comparisons between Hippocampal Volumetry, Tensor-Based Morphometry and Voxel-Based Morphometry

PubMed Central

Muñoz-Ruiz, Miguel Ángel; Hartikainen, Päivi; Koikkalainen, Juha; Wolz, Robin; Julkunen, Valtteri; Niskanen, Eini; Herukka, Sanna-Kaisa; Kivipelto, Miia; Vanninen, Ritva; Rueckert, Daniel; Liu, Yawu; Lötjönen, Jyrki; Soininen, Hilkka

2012-01-01

Background MRI is an important clinical tool for diagnosing dementia-like diseases such as Frontemporal Dementia (FTD). However there is a need to develop more accurate and standardized MRI analysis methods. Objective To compare FTD with Alzheimer’s Disease (AD) and Mild Cognitive Impairment (MCI) with three automatic MRI analysis methods - Hippocampal Volumetry (HV), Tensor-based Morphometry (TBM) and Voxel-based Morphometry (VBM), in specific regions of interest in order to determine the highest classification accuracy. Methods Thirty-seven patients with FTD, 46 patients with AD, 26 control subjects, 16 patients with progressive MCI (PMCI) and 48 patients with stable MCI (SMCI) were examined with HV, TBM for shape change, and VBM for gray matter density. We calculated the Correct Classification Rate (CCR), sensitivity (SS) and specificity (SP) between the study groups. Results We found unequivocal results differentiating controls from FTD with HV (hippocampus left side) (CCR = 0.83; SS = 0.84; SP = 0.80), with TBM (hippocampus and amygdala (CCR = 0.80/SS = 0.71/SP = 0.94), and with VBM (all the regions studied, especially in lateral ventricle frontal horn, central part and occipital horn) (CCR = 0.87/SS = 0.81/SP = 0.96). VBM achieved the highest accuracy in differentiating AD and FTD (CCR = 0.72/SS = 0.67/SP = 0.76), particularly in lateral ventricle (frontal horn, central part and occipital horn) (CCR = 0.73), whereas TBM in superior frontal gyrus also achieved a high accuracy (CCR = 0.71/SS = 0.68/SP = 0.73). TBM resulted in low accuracy (CCR = 0.62) in the differentiation of AD from FTD using all regions of interest, with similar results for HV (CCR = 0.55). Conclusion Hippocampal atrophy is present not only in AD but also in FTD. Of the methods used, VBM achieved the highest accuracy in its ability to differentiate between FTD and AD. PMID:23285078

Design of Embedded System for Multivariate Classification of Finger and Thumb Movements Using EEG Signals for Control of Upper Limb Prosthesis.

PubMed

Rashid, Nasir; Iqbal, Javaid; Javed, Amna; Tiwana, Mohsin I; Khan, Umar Shahbaz

2018-01-01

Brain Computer Interface (BCI) determines the intent of the user from a variety of electrophysiological signals. These signals, Slow Cortical Potentials, are recorded from scalp, and cortical neuronal activity is recorded by implanted electrodes. This paper is focused on design of an embedded system that is used to control the finger movements of an upper limb prosthesis using Electroencephalogram (EEG) signals. This is a follow-up of our previous research which explored the best method to classify three movements of fingers (thumb movement, index finger movement, and first movement). Two-stage logistic regression classifier exhibited the highest classification accuracy while Power Spectral Density (PSD) was used as a feature of the filtered signal. The EEG signal data set was recorded using a 14-channel electrode headset (a noninvasive BCI system) from right-handed, neurologically intact volunteers. Mu (commonly known as alpha waves) and Beta Rhythms (8-30 Hz) containing most of the movement data were retained through filtering using "Arduino Uno" microcontroller followed by 2-stage logistic regression to obtain a mean classification accuracy of 70%.

An Activity Recognition Framework Deploying the Random Forest Classifier and A Single Optical Heart Rate Monitoring and Triaxial AccelerometerWrist-Band.

PubMed

Mehrang, Saeed; Pietilä, Julia; Korhonen, Ilkka

2018-02-22

Wrist-worn sensors have better compliance for activity monitoring compared to hip, waist, ankle or chest positions. However, wrist-worn activity monitoring is challenging due to the wide degree of freedom for the hand movements, as well as similarity of hand movements in different activities such as varying intensities of cycling. To strengthen the ability of wrist-worn sensors in detecting human activities more accurately, motion signals can be complemented by physiological signals such as optical heart rate (HR) based on photoplethysmography. In this paper, an activity monitoring framework using an optical HR sensor and a triaxial wrist-worn accelerometer is presented. We investigated a range of daily life activities including sitting, standing, household activities and stationary cycling with two intensities. A random forest (RF) classifier was exploited to detect these activities based on the wrist motions and optical HR. The highest overall accuracy of 89.6 ± 3.9% was achieved with a forest of a size of 64 trees and 13-s signal segments with 90% overlap. Removing the HR-derived features decreased the classification accuracy of high-intensity cycling by almost 7%, but did not affect the classification accuracies of other activities. A feature reduction utilizing the feature importance scores of RF was also carried out and resulted in a shrunken feature set of only 21 features. The overall accuracy of the classification utilizing the shrunken feature set was 89.4 ± 4.2%, which is almost equivalent to the above-mentioned peak overall accuracy.

Comparison of Random Forest, k-Nearest Neighbor, and Support Vector Machine Classifiers for Land Cover Classification Using Sentinel-2 Imagery

PubMed Central

Thanh Noi, Phan; Kappas, Martin

2017-01-01

In previous classification studies, three non-parametric classifiers, Random Forest (RF), k-Nearest Neighbor (kNN), and Support Vector Machine (SVM), were reported as the foremost classifiers at producing high accuracies. However, only a few studies have compared the performances of these classifiers with different training sample sizes for the same remote sensing images, particularly the Sentinel-2 Multispectral Imager (MSI). In this study, we examined and compared the performances of the RF, kNN, and SVM classifiers for land use/cover classification using Sentinel-2 image data. An area of 30 × 30 km2 within the Red River Delta of Vietnam with six land use/cover types was classified using 14 different training sample sizes, including balanced and imbalanced, from 50 to over 1250 pixels/class. All classification results showed a high overall accuracy (OA) ranging from 90% to 95%. Among the three classifiers and 14 sub-datasets, SVM produced the highest OA with the least sensitivity to the training sample sizes, followed consecutively by RF and kNN. In relation to the sample size, all three classifiers showed a similar and high OA (over 93.85%) when the training sample size was large enough, i.e., greater than 750 pixels/class or representing an area of approximately 0.25% of the total study area. The high accuracy was achieved with both imbalanced and balanced datasets. PMID:29271909

Comparison of Random Forest, k-Nearest Neighbor, and Support Vector Machine Classifiers for Land Cover Classification Using Sentinel-2 Imagery.

PubMed

Thanh Noi, Phan; Kappas, Martin

2017-12-22

In previous classification studies, three non-parametric classifiers, Random Forest (RF), k-Nearest Neighbor (kNN), and Support Vector Machine (SVM), were reported as the foremost classifiers at producing high accuracies. However, only a few studies have compared the performances of these classifiers with different training sample sizes for the same remote sensing images, particularly the Sentinel-2 Multispectral Imager (MSI). In this study, we examined and compared the performances of the RF, kNN, and SVM classifiers for land use/cover classification using Sentinel-2 image data. An area of 30 × 30 km² within the Red River Delta of Vietnam with six land use/cover types was classified using 14 different training sample sizes, including balanced and imbalanced, from 50 to over 1250 pixels/class. All classification results showed a high overall accuracy (OA) ranging from 90% to 95%. Among the three classifiers and 14 sub-datasets, SVM produced the highest OA with the least sensitivity to the training sample sizes, followed consecutively by RF and kNN. In relation to the sample size, all three classifiers showed a similar and high OA (over 93.85%) when the training sample size was large enough, i.e., greater than 750 pixels/class or representing an area of approximately 0.25% of the total study area. The high accuracy was achieved with both imbalanced and balanced datasets.

Ranking of predictor variables based on effect size criterion provides an accurate means of automatically classifying opinion column articles

NASA Astrophysics Data System (ADS)

Legara, Erika Fille; Monterola, Christopher; Abundo, Cheryl

2011-01-01

We demonstrate an accurate procedure based on linear discriminant analysis that allows automatic authorship classification of opinion column articles. First, we extract the following stylometric features of 157 column articles from four authors: statistics on high frequency words, number of words per sentence, and number of sentences per paragraph. Then, by systematically ranking these features based on an effect size criterion, we show that we can achieve an average classification accuracy of 93% for the test set. In comparison, frequency size based ranking has an average accuracy of 80%. The highest possible average classification accuracy of our data merely relying on chance is ∼31%. By carrying out sensitivity analysis, we show that the effect size criterion is superior than frequency ranking because there exist low frequency words that significantly contribute to successful author discrimination. Consistent results are seen when the procedure is applied in classifying the undisputed Federalist papers of Alexander Hamilton and James Madison. To the best of our knowledge, the work is the first attempt in classifying opinion column articles, that by virtue of being shorter in length (as compared to novels or short stories), are more prone to over-fitting issues. The near perfect classification for the longer papers supports this claim. Our results provide an important insight on authorship attribution that has been overlooked in previous studies: that ranking discriminant variables based on word frequency counts is not necessarily an optimal procedure.

Heritage landscape structure analysis in surrounding environment of the Grand Canal Yangzhou section

NASA Astrophysics Data System (ADS)

Xu, Huan

2018-03-01

The Yangzhou section of the Grand Canal is selected for a case study in this paper. The ZY-3 satellite images of 2016 are adopted as the data source. RS and GIS are used to analyze the landscape classification of the surrounding landscape of the Grand Canal, and the classification results are precisely evaluated. Next, the overall features of the landscape pattern are analyzed. The results showed that the overall accuracy is 82.5% and the Kappa coefficient is 78.17% in the Yangzhou section. The producer’s accuracy of the water landscape is the highest, followed by that of the other landscape, farmland landscape, garden and forest landscape, architectural landscape. The user’s accuracy of different landscape types can be ranked in a descending order, as the water landscape, farmland landscape, road landscape, architectural landscape, other landscape, garden and forest landscape. The farmland landscape and the architectural landscape are the top advantageous landscape types of the heritage site. The research findings can provide basic data for landscape protection, management and sustainable development of the Grand Canal Yangzhou section.

The dynamics of human-induced land cover change in miombo ecosystems of southern Africa

NASA Astrophysics Data System (ADS)

Jaiteh, Malanding Sambou

Understanding human-induced land cover change in the miombo require the consistent, geographically-referenced, data on temporal land cover characteristics as well as biophysical and socioeconomic drivers of land use, the major cause of land cover change. The overall goal of this research to examine the applications of high-resolution satellite remote sensing data in studying the dynamics of human-induced land cover change in the miombo. Specific objectives are to: (1) evaluate the applications of computer-assisted classification of Landsat Thematic Mapper (TM) data for land cover mapping in the miombo and (2) analyze spatial and temporal patterns of landscape change locations in the miombo. Stepwise Thematic Classification, STC (a hybrid supervised-unsupervised classification) procedure for classifying Landsat TM data was developed and tested using Landsat TM data. Classification accuracy results were compared to those from supervised and unsupervised classification. The STC provided the highest classification accuracy i.e., 83.9% correspondence between classified and referenced data compared to 44.2% and 34.5% for unsupervised and supervised classification respectively. Improvements in the classification process can be attributed to thematic stratification of the image data into spectrally homogenous (thematic) groups and step-by-step classification of the groups using supervised or unsupervised classification techniques. Supervised classification failed to classify 18% of the scene evidence that training data used did not adequately represent all of the variability in the data. Application of the procedure in drier miombo produced overall classification accuracy of 63%. This is much lower than that of wetter miombo. The results clearly demonstrate that digital classification of Landsat TM can be successfully implemented in the miombo without intensive fieldwork. Spatial characteristics of land cover change in agricultural and forested landscapes in central Malawi were analyzed for the period 1984 to 1995 spatial pattern analysis methods. Shifting cultivation areas, Agriculture in forested landscape, experienced highest rate of woodland cover fragmentation with mean patch size of closed woodland cover decreasing from 20ha to 7.5ha. Permanent bare (cropland and settlement) in intensive agricultural matrix landscapes increased 52% largely through the conversion of fallow areas. Protected National Park area remained fairly unchanged although closed woodland area increased by 4%, mainly from regeneration of open woodland. This study provided evidence that changes in spatial characteristics in the miombo differ with landscape. Land use change (i.e. conversion to cropland) is the primary driving force behind changes in landscape spatial patterns. Also, results revealed that exclusion of intense human use (i.e. cultivation and woodcutting) through regulations and/or fencing increased both closed woodland area (through regeneration of open woodland) and overall connectivity in the landscape. Spatial characteristics of land cover change were analyzed at locations in Malawi (wetter miombo) and Zimbabwe (drier miombo). Results indicate land cover dynamics differ both between and within case study sites. In communal areas in the Kasungu scene, land cover change is dominated by woodland fragmentation to open vegetation. Change in private commercial lands was dominantly expansion of bare (settlement and cropland) areas primarily at the expense of open vegetation (fallow land).

Parameterization of Shape and Compactness in Object-based Image Classification Using Quickbird-2 Imagery

NASA Astrophysics Data System (ADS)

Tonbul, H.; Kavzoglu, T.

2016-12-01

In recent years, object based image analysis (OBIA) has spread out and become a widely accepted technique for the analysis of remotely sensed data. OBIA deals with grouping pixels into homogenous objects based on spectral, spatial and textural features of contiguous pixels in an image. The first stage of OBIA, named as image segmentation, is the most prominent part of object recognition. In this study, multiresolution segmentation, which is a region-based approach, was employed to construct image objects. In the application of multi-resolution, three parameters, namely shape, compactness and scale must be set by the analyst. Segmentation quality remarkably influences the fidelity of the thematic maps and accordingly the classification accuracy. Therefore, it is of great importance to search and set optimal values for the segmentation parameters. In the literature, main focus has been on the definition of scale parameter, assuming that the effect of shape and compactness parameters is limited in terms of achieved classification accuracy. The aim of this study is to deeply analyze the influence of shape/compactness parameters by varying their values while using the optimal scale parameter determined by the use of Estimation of Scale Parameter (ESP-2) approach. A pansharpened Qickbird-2 image covering Trabzon, Turkey was employed to investigate the objectives of the study. For this purpose, six different combinations of shape/compactness were utilized to make deductions on the behavior of shape and compactness parameters and optimal setting for all parameters as a whole. Objects were assigned to classes using nearest neighbor classifier in all segmentation observations and equal number of pixels was randomly selected to calculate accuracy metrics. The highest overall accuracy (92.3%) was achieved by setting the shape/compactness criteria to 0.3/0.3. The results of this study indicate that shape/compactness parameters can have significant effect on classification accuracy with 4% change in overall accuracy. Also, statistical significance of differences in accuracy was tested using the McNemar's test and found that the difference between poor and optimal setting of shape/compactness parameters was statistically significant, suggesting a search for optimal parameterization instead of default setting.

Mapping permafrost in the boreal forest with Thematic Mapper satellite data

NASA Technical Reports Server (NTRS)

Morrissey, L. A.; Strong, L. L.; Card, D. H.

1986-01-01

A geographic data base incorporating Landsat TM data was used to develop and evaluate logistic discriminant functions for predicting the distribution of permafrost in a boreal forest watershed. The data base included both satellite-derived information and ancillary map data. Five permafrost classifications were developed from a stratified random sample of the data base and evaluated by comparison with a photo-interpreted permafrost map using contingency table analysis and soil temperatures recorded at sites within the watershed. A classification using a TM thermal band and a TM-derived vegetation map as independent variables yielded the highest mapping accuracy for all permafrost categories.

A new hybrid method based on fuzzy-artificial immune system and k-nn algorithm for breast cancer diagnosis.

PubMed

Sahan, Seral; Polat, Kemal; Kodaz, Halife; Güneş, Salih

2007-03-01

The use of machine learning tools in medical diagnosis is increasing gradually. This is mainly because the effectiveness of classification and recognition systems has improved in a great deal to help medical experts in diagnosing diseases. Such a disease is breast cancer, which is a very common type of cancer among woman. As the incidence of this disease has increased significantly in the recent years, machine learning applications to this problem have also took a great attention as well as medical consideration. This study aims at diagnosing breast cancer with a new hybrid machine learning method. By hybridizing a fuzzy-artificial immune system with k-nearest neighbour algorithm, a method was obtained to solve this diagnosis problem via classifying Wisconsin Breast Cancer Dataset (WBCD). This data set is a very commonly used data set in the literature relating the use of classification systems for breast cancer diagnosis and it was used in this study to compare the classification performance of our proposed method with regard to other studies. We obtained a classification accuracy of 99.14%, which is the highest one reached so far. The classification accuracy was obtained via 10-fold cross validation. This result is for WBCD but it states that this method can be used confidently for other breast cancer diagnosis problems, too.

Detection of epileptic seizure in EEG signals using linear least squares preprocessing.

PubMed

Roshan Zamir, Z

2016-09-01

An epileptic seizure is a transient event of abnormal excessive neuronal discharge in the brain. This unwanted event can be obstructed by detection of electrical changes in the brain that happen before the seizure takes place. The automatic detection of seizures is necessary since the visual screening of EEG recordings is a time consuming task and requires experts to improve the diagnosis. Much of the prior research in detection of seizures has been developed based on artificial neural network, genetic programming, and wavelet transforms. Although the highest achieved accuracy for classification is 100%, there are drawbacks, such as the existence of unbalanced datasets and the lack of investigations in performances consistency. To address these, four linear least squares-based preprocessing models are proposed to extract key features of an EEG signal in order to detect seizures. The first two models are newly developed. The original signal (EEG) is approximated by a sinusoidal curve. Its amplitude is formed by a polynomial function and compared with the predeveloped spline function. Different statistical measures, namely classification accuracy, true positive and negative rates, false positive and negative rates and precision, are utilised to assess the performance of the proposed models. These metrics are derived from confusion matrices obtained from classifiers. Different classifiers are used over the original dataset and the set of extracted features. The proposed models significantly reduce the dimension of the classification problem and the computational time while the classification accuracy is improved in most cases. The first and third models are promising feature extraction methods with the classification accuracy of 100%. Logistic, LazyIB1, LazyIB5, and J48 are the best classifiers. Their true positive and negative rates are 1 while false positive and negative rates are 0 and the corresponding precision values are 1. Numerical results suggest that these models are robust and efficient for detecting epileptic seizure. Copyright © 2016 Elsevier Ireland Ltd. All rights reserved.

An Evaluation of Item Response Theory Classification Accuracy and Consistency Indices

ERIC Educational Resources Information Center

Wyse, Adam E.; Hao, Shiqi

2012-01-01

This article introduces two new classification consistency indices that can be used when item response theory (IRT) models have been applied. The new indices are shown to be related to Rudner's classification accuracy index and Guo's classification accuracy index. The Rudner- and Guo-based classification accuracy and consistency indices are…

Identification of suitable genes contributes to lung adenocarcinoma clustering by multiple meta-analysis methods.

PubMed

Yang, Ze-Hui; Zheng, Rui; Gao, Yuan; Zhang, Qiang

2016-09-01

With the widespread application of high-throughput technology, numerous meta-analysis methods have been proposed for differential expression profiling across multiple studies. We identified the suitable differentially expressed (DE) genes that contributed to lung adenocarcinoma (ADC) clustering based on seven popular multiple meta-analysis methods. Seven microarray expression profiles of ADC and normal controls were extracted from the ArrayExpress database. The Bioconductor was used to perform the data preliminary preprocessing. Then, DE genes across multiple studies were identified. Hierarchical clustering was applied to compare the classification performance for microarray data samples. The classification efficiency was compared based on accuracy, sensitivity and specificity. Across seven datasets, 573 ADC cases and 222 normal controls were collected. After filtering out unexpressed and noninformative genes, 3688 genes were remained for further analysis. The classification efficiency analysis showed that DE genes identified by sum of ranks method separated ADC from normal controls with the best accuracy, sensitivity and specificity of 0.953, 0.969 and 0.932, respectively. The gene set with the highest classification accuracy mainly participated in the regulation of response to external stimulus (P = 7.97E-04), cyclic nucleotide-mediated signaling (P = 0.01), regulation of cell morphogenesis (P = 0.01) and regulation of cell proliferation (P = 0.01). Evaluation of DE genes identified by different meta-analysis methods in classification efficiency provided a new perspective to the choice of the suitable method in a given application. Varying meta-analysis methods always present varying abilities, so synthetic consideration should be taken when providing meta-analysis methods for particular research. © 2015 John Wiley & Sons Ltd.

Intelligent feature selection techniques for pattern classification of Lamb wave signals

DOE Office of Scientific and Technical Information (OSTI.GOV)

Hinders, Mark K.; Miller, Corey A.

2014-02-18

Lamb wave interaction with flaws is a complex, three-dimensional phenomenon, which often frustrates signal interpretation schemes based on mode arrival time shifts predicted by dispersion curves. As the flaw severity increases, scattering and mode conversion effects will often dominate the time-domain signals, obscuring available information about flaws because multiple modes may arrive on top of each other. Even for idealized flaw geometries the scattering and mode conversion behavior of Lamb waves is very complex. Here, multi-mode Lamb waves in a metal plate are propagated across a rectangular flat-bottom hole in a sequence of pitch-catch measurements corresponding to the double crossholemore » tomography geometry. The flaw is sequentially deepened, with the Lamb wave measurements repeated at each flaw depth. Lamb wave tomography reconstructions are used to identify which waveforms have interacted with the flaw and thereby carry information about its depth. Multiple features are extracted from each of the Lamb wave signals using wavelets, which are then fed to statistical pattern classification algorithms that identify flaw severity. In order to achieve the highest classification accuracy, an optimal feature space is required but it’s never known a priori which features are going to be best. For structural health monitoring we make use of the fact that physical flaws, such as corrosion, will only increase over time. This allows us to identify feature vectors which are topologically well-behaved by requiring that sequential classes “line up” in feature vector space. An intelligent feature selection routine is illustrated that identifies favorable class distributions in multi-dimensional feature spaces using computational homology theory. Betti numbers and formal classification accuracies are calculated for each feature space subset to establish a correlation between the topology of the class distribution and the corresponding classification accuracy.« less

Assessment of pedophilia using hemodynamic brain response to sexual stimuli.

PubMed

Ponseti, Jorge; Granert, Oliver; Jansen, Olav; Wolff, Stephan; Beier, Klaus; Neutze, Janina; Deuschl, Günther; Mehdorn, Hubertus; Siebner, Hartwig; Bosinski, Hartmut

2012-02-01

Accurately assessing sexual preference is important in the treatment of child sex offenders. Phallometry is the standard method to identify sexual preference; however, this measure has been criticized for its intrusiveness and limited reliability. To evaluate whether spatial response pattern to sexual stimuli as revealed by a change in the blood oxygen level-dependent signal facilitates the identification of pedophiles. During functional magnetic resonance imaging, pedophilic and nonpedophilic participants were briefly exposed to same- and opposite-sex images of nude children and adults. We calculated differences in blood oxygen level-dependent signals to child and adult sexual stimuli for each participant. The corresponding contrast images were entered into a group analysis to calculate whole-brain difference maps between groups. We calculated an expression value that corresponded to the group result for each participant. These expression values were submitted to 2 different classification algorithms: Fisher linear discriminant analysis and κ -nearest neighbor analysis. This classification procedure was cross-validated using the leave-one-out method. Section of Sexual Medicine, Medical School, Christian Albrechts University of Kiel, Kiel, Germany. We recruited 24 participants with pedophilia who were sexually attracted to either prepubescent girls (n = 11) or prepubescent boys (n = 13) and 32 healthy male controls who were sexually attracted to either adult women (n = 18) or adult men (n = 14). Sensitivity and specificity scores of the 2 classification algorithms. The highest classification accuracy was achieved by Fisher linear discriminant analysis, which showed a mean accuracy of 95% (100% specificity, 88% sensitivity). Functional brain response patterns to sexual stimuli contain sufficient information to identify pedophiles with high accuracy. The automatic classification of these patterns is a promising objective tool to clinically diagnose pedophilia.

A Comparative Analysis of Machine Learning with WorldView-2 Pan-Sharpened Imagery for Tea Crop Mapping

PubMed Central

Chuang, Yung-Chung Matt; Shiu, Yi-Shiang

2016-01-01

Tea is an important but vulnerable economic crop in East Asia, highly impacted by climate change. This study attempts to interpret tea land use/land cover (LULC) using very high resolution WorldView-2 imagery of central Taiwan with both pixel and object-based approaches. A total of 80 variables derived from each WorldView-2 band with pan-sharpening, standardization, principal components and gray level co-occurrence matrix (GLCM) texture indices transformation, were set as the input variables. For pixel-based image analysis (PBIA), 34 variables were selected, including seven principal components, 21 GLCM texture indices and six original WorldView-2 bands. Results showed that support vector machine (SVM) had the highest tea crop classification accuracy (OA = 84.70% and KIA = 0.690), followed by random forest (RF), maximum likelihood algorithm (ML), and logistic regression analysis (LR). However, the ML classifier achieved the highest classification accuracy (OA = 96.04% and KIA = 0.887) in object-based image analysis (OBIA) using only six variables. The contribution of this study is to create a new framework for accurately identifying tea crops in a subtropical region with real-time high-resolution WorldView-2 imagery without field survey, which could further aid agriculture land management and a sustainable agricultural product supply. PMID:27128915

A Comparative Analysis of Machine Learning with WorldView-2 Pan-Sharpened Imagery for Tea Crop Mapping.

PubMed

Chuang, Yung-Chung Matt; Shiu, Yi-Shiang

2016-04-26

Tea is an important but vulnerable economic crop in East Asia, highly impacted by climate change. This study attempts to interpret tea land use/land cover (LULC) using very high resolution WorldView-2 imagery of central Taiwan with both pixel and object-based approaches. A total of 80 variables derived from each WorldView-2 band with pan-sharpening, standardization, principal components and gray level co-occurrence matrix (GLCM) texture indices transformation, were set as the input variables. For pixel-based image analysis (PBIA), 34 variables were selected, including seven principal components, 21 GLCM texture indices and six original WorldView-2 bands. Results showed that support vector machine (SVM) had the highest tea crop classification accuracy (OA = 84.70% and KIA = 0.690), followed by random forest (RF), maximum likelihood algorithm (ML), and logistic regression analysis (LR). However, the ML classifier achieved the highest classification accuracy (OA = 96.04% and KIA = 0.887) in object-based image analysis (OBIA) using only six variables. The contribution of this study is to create a new framework for accurately identifying tea crops in a subtropical region with real-time high-resolution WorldView-2 imagery without field survey, which could further aid agriculture land management and a sustainable agricultural product supply.

Capability of Integrated MODIS Imagery and ALOS for Oil Palm, Rubber and Forest Areas Mapping in Tropical Forest Regions

PubMed Central

Razali, Sheriza Mohd; Marin, Arnaldo; Nuruddin, Ahmad Ainuddin; Shafri, Helmi Zulhaidi Mohd; Hamid, Hazandy Abdul

2014-01-01

Various classification methods have been applied for low resolution of the entire Earth's surface from recorded satellite images, but insufficient study has determined which method, for which satellite data, is economically viable for tropical forest land use mapping. This study employed Iterative Self Organizing Data Analysis Techniques (ISODATA) and K-Means classification techniques to classified Moderate Resolution Imaging Spectroradiometer (MODIS) Surface Reflectance satellite image into forests, oil palm groves, rubber plantations, mixed horticulture, mixed oil palm and rubber and mixed forest and rubber. Even though frequent cloud cover has been a challenge for mapping tropical forests, our MODIS land use classification map found that 2008 ISODATA-1 performed well with overall accuracy of 94%, with the highest Producer's Accuracy of Forest with 86%, and were consistent with MODIS Land Cover 2008 (MOD12Q1), respectively. The MODIS land use classification was able to distinguish young oil palm groves from open areas, rubber and mature oil palm plantations, on the Advanced Land Observing Satellite (ALOS) map, whereas rubber was more easily distinguished from an open area than from mixed rubber and forest. This study provides insight on the potential for integrating regional databases and temporal MODIS data, in order to map land use in tropical forest regions. PMID:24811079

Capability of integrated MODIS imagery and ALOS for oil palm, rubber and forest areas mapping in tropical forest regions.

PubMed

Razali, Sheriza Mohd; Marin, Arnaldo; Nuruddin, Ahmad Ainuddin; Shafri, Helmi Zulhaidi Mohd; Hamid, Hazandy Abdul

2014-05-07

Various classification methods have been applied for low resolution of the entire Earth's surface from recorded satellite images, but insufficient study has determined which method, for which satellite data, is economically viable for tropical forest land use mapping. This study employed Iterative Self Organizing Data Analysis Techniques (ISODATA) and K-Means classification techniques to classified Moderate Resolution Imaging Spectroradiometer (MODIS) Surface Reflectance satellite image into forests, oil palm groves, rubber plantations, mixed horticulture, mixed oil palm and rubber and mixed forest and rubber. Even though frequent cloud cover has been a challenge for mapping tropical forests, our MODIS land use classification map found that 2008 ISODATA-1 performed well with overall accuracy of 94%, with the highest Producer's Accuracy of Forest with 86%, and were consistent with MODIS Land Cover 2008 (MOD12Q1), respectively. The MODIS land use classification was able to distinguish young oil palm groves from open areas, rubber and mature oil palm plantations, on the Advanced Land Observing Satellite (ALOS) map, whereas rubber was more easily distinguished from an open area than from mixed rubber and forest. This study provides insight on the potential for integrating regional databases and temporal MODIS data, in order to map land use in tropical forest regions.

A contour-based shape descriptor for biomedical image classification and retrieval

NASA Astrophysics Data System (ADS)

You, Daekeun; Antani, Sameer; Demner-Fushman, Dina; Thoma, George R.

2013-12-01

Contours, object blobs, and specific feature points are utilized to represent object shapes and extract shape descriptors that can then be used for object detection or image classification. In this research we develop a shape descriptor for biomedical image type (or, modality) classification. We adapt a feature extraction method used in optical character recognition (OCR) for character shape representation, and apply various image preprocessing methods to successfully adapt the method to our application. The proposed shape descriptor is applied to radiology images (e.g., MRI, CT, ultrasound, X-ray, etc.) to assess its usefulness for modality classification. In our experiment we compare our method with other visual descriptors such as CEDD, CLD, Tamura, and PHOG that extract color, texture, or shape information from images. The proposed method achieved the highest classification accuracy of 74.1% among all other individual descriptors in the test, and when combined with CSD (color structure descriptor) showed better performance (78.9%) than using the shape descriptor alone.

Histogram Curve Matching Approaches for Object-based Image Classification of Land Cover and Land Use

PubMed Central

Toure, Sory I.; Stow, Douglas A.; Weeks, John R.; Kumar, Sunil

2013-01-01

The classification of image-objects is usually done using parametric statistical measures of central tendency and/or dispersion (e.g., mean or standard deviation). The objectives of this study were to analyze digital number histograms of image objects and evaluate classifications measures exploiting characteristic signatures of such histograms. Two histograms matching classifiers were evaluated and compared to the standard nearest neighbor to mean classifier. An ADS40 airborne multispectral image of San Diego, California was used for assessing the utility of curve matching classifiers in a geographic object-based image analysis (GEOBIA) approach. The classifications were performed with data sets having 0.5 m, 2.5 m, and 5 m spatial resolutions. Results show that histograms are reliable features for characterizing classes. Also, both histogram matching classifiers consistently performed better than the one based on the standard nearest neighbor to mean rule. The highest classification accuracies were produced with images having 2.5 m spatial resolution. PMID:24403648

Multivariate Classification of Major Depressive Disorder Using the Effective Connectivity and Functional Connectivity

PubMed Central

Geng, Xiangfei; Xu, Junhai; Liu, Baolin; Shi, Yonggang

2018-01-01

Major depressive disorder (MDD) is a mental disorder characterized by at least 2 weeks of low mood, which is present across most situations. Diagnosis of MDD using rest-state functional magnetic resonance imaging (fMRI) data faces many challenges due to the high dimensionality, small samples, noisy and individual variability. To our best knowledge, no studies aim at classification with effective connectivity and functional connectivity measures between MDD patients and healthy controls. In this study, we performed a data-driving classification analysis using the whole brain connectivity measures which included the functional connectivity from two brain templates and effective connectivity measures created by the default mode network (DMN), dorsal attention network (DAN), frontal-parietal network (FPN), and silence network (SN). Effective connectivity measures were extracted using spectral Dynamic Causal Modeling (spDCM) and transformed into a vectorial feature space. Linear Support Vector Machine (linear SVM), non-linear SVM, k-Nearest Neighbor (KNN), and Logistic Regression (LR) were used as the classifiers to identify the differences between MDD patients and healthy controls. Our results showed that the highest accuracy achieved 91.67% (p < 0.0001) when using 19 effective connections and 89.36% when using 6,650 functional connections. The functional connections with high discriminative power were mainly located within or across the whole brain resting-state networks while the discriminative effective connections located in several specific regions, such as posterior cingulate cortex (PCC), ventromedial prefrontal cortex (vmPFC), dorsal cingulate cortex (dACC), and inferior parietal lobes (IPL). To further compare the discriminative power of functional connections and effective connections, a classification analysis only using the functional connections from those four networks was conducted and the highest accuracy achieved 78.33% (p < 0.0001). Our study demonstrated that the effective connectivity measures might play a more important role than functional connectivity in exploring the alterations between patients and health controls and afford a better mechanistic interpretability. Moreover, our results showed a diagnostic potential of the effective connectivity for the diagnosis of MDD patients with high accuracies allowing for earlier prevention or intervention. PMID:29515348

Deep Learning Accurately Predicts Estrogen Receptor Status in Breast Cancer Metabolomics Data.

PubMed

Alakwaa, Fadhl M; Chaudhary, Kumardeep; Garmire, Lana X

2018-01-05

Metabolomics holds the promise as a new technology to diagnose highly heterogeneous diseases. Conventionally, metabolomics data analysis for diagnosis is done using various statistical and machine learning based classification methods. However, it remains unknown if deep neural network, a class of increasingly popular machine learning methods, is suitable to classify metabolomics data. Here we use a cohort of 271 breast cancer tissues, 204 positive estrogen receptor (ER+), and 67 negative estrogen receptor (ER-) to test the accuracies of feed-forward networks, a deep learning (DL) framework, as well as six widely used machine learning models, namely random forest (RF), support vector machines (SVM), recursive partitioning and regression trees (RPART), linear discriminant analysis (LDA), prediction analysis for microarrays (PAM), and generalized boosted models (GBM). DL framework has the highest area under the curve (AUC) of 0.93 in classifying ER+/ER- patients, compared to the other six machine learning algorithms. Furthermore, the biological interpretation of the first hidden layer reveals eight commonly enriched significant metabolomics pathways (adjusted P-value <0.05) that cannot be discovered by other machine learning methods. Among them, protein digestion and absorption and ATP-binding cassette (ABC) transporters pathways are also confirmed in integrated analysis between metabolomics and gene expression data in these samples. In summary, deep learning method shows advantages for metabolomics based breast cancer ER status classification, with both the highest prediction accuracy (AUC = 0.93) and better revelation of disease biology. We encourage the adoption of feed-forward networks based deep learning method in the metabolomics research community for classification.

An efficient diagnosis system for Parkinson's disease using kernel-based extreme learning machine with subtractive clustering features weighting approach.

PubMed

Ma, Chao; Ouyang, Jihong; Chen, Hui-Ling; Zhao, Xue-Hua

2014-01-01

A novel hybrid method named SCFW-KELM, which integrates effective subtractive clustering features weighting and a fast classifier kernel-based extreme learning machine (KELM), has been introduced for the diagnosis of PD. In the proposed method, SCFW is used as a data preprocessing tool, which aims at decreasing the variance in features of the PD dataset, in order to further improve the diagnostic accuracy of the KELM classifier. The impact of the type of kernel functions on the performance of KELM has been investigated in detail. The efficiency and effectiveness of the proposed method have been rigorously evaluated against the PD dataset in terms of classification accuracy, sensitivity, specificity, area under the receiver operating characteristic (ROC) curve (AUC), f-measure, and kappa statistics value. Experimental results have demonstrated that the proposed SCFW-KELM significantly outperforms SVM-based, KNN-based, and ELM-based approaches and other methods in the literature and achieved highest classification results reported so far via 10-fold cross validation scheme, with the classification accuracy of 99.49%, the sensitivity of 100%, the specificity of 99.39%, AUC of 99.69%, the f-measure value of 0.9964, and kappa value of 0.9867. Promisingly, the proposed method might serve as a new candidate of powerful methods for the diagnosis of PD with excellent performance.

An Efficient Diagnosis System for Parkinson's Disease Using Kernel-Based Extreme Learning Machine with Subtractive Clustering Features Weighting Approach

PubMed Central

Ma, Chao; Ouyang, Jihong; Chen, Hui-Ling; Zhao, Xue-Hua

2014-01-01

A novel hybrid method named SCFW-KELM, which integrates effective subtractive clustering features weighting and a fast classifier kernel-based extreme learning machine (KELM), has been introduced for the diagnosis of PD. In the proposed method, SCFW is used as a data preprocessing tool, which aims at decreasing the variance in features of the PD dataset, in order to further improve the diagnostic accuracy of the KELM classifier. The impact of the type of kernel functions on the performance of KELM has been investigated in detail. The efficiency and effectiveness of the proposed method have been rigorously evaluated against the PD dataset in terms of classification accuracy, sensitivity, specificity, area under the receiver operating characteristic (ROC) curve (AUC), f-measure, and kappa statistics value. Experimental results have demonstrated that the proposed SCFW-KELM significantly outperforms SVM-based, KNN-based, and ELM-based approaches and other methods in the literature and achieved highest classification results reported so far via 10-fold cross validation scheme, with the classification accuracy of 99.49%, the sensitivity of 100%, the specificity of 99.39%, AUC of 99.69%, the f-measure value of 0.9964, and kappa value of 0.9867. Promisingly, the proposed method might serve as a new candidate of powerful methods for the diagnosis of PD with excellent performance. PMID:25484912

Classification of Suicide Attempts through a Machine Learning Algorithm Based on Multiple Systemic Psychiatric Scales.

PubMed

Oh, Jihoon; Yun, Kyongsik; Hwang, Ji-Hyun; Chae, Jeong-Ho

2017-01-01

Classification and prediction of suicide attempts in high-risk groups is important for preventing suicide. The purpose of this study was to investigate whether the information from multiple clinical scales has classification power for identifying actual suicide attempts. Patients with depression and anxiety disorders ( N  = 573) were included, and each participant completed 31 self-report psychiatric scales and questionnaires about their history of suicide attempts. We then trained an artificial neural network classifier with 41 variables (31 psychiatric scales and 10 sociodemographic elements) and ranked the contribution of each variable for the classification of suicide attempts. To evaluate the clinical applicability of our model, we measured classification performance with top-ranked predictors. Our model had an overall accuracy of 93.7% in 1-month, 90.8% in 1-year, and 87.4% in lifetime suicide attempts detection. The area under the receiver operating characteristic curve (AUROC) was the highest for 1-month suicide attempts detection (0.93), followed by lifetime (0.89), and 1-year detection (0.87). Among all variables, the Emotion Regulation Questionnaire had the highest contribution, and the positive and negative characteristics of the scales similarly contributed to classification performance. Performance on suicide attempts classification was largely maintained when we only used the top five ranked variables for training (AUROC; 1-month, 0.75, 1-year, 0.85, lifetime suicide attempts detection, 0.87). Our findings indicate that information from self-report clinical scales can be useful for the classification of suicide attempts. Based on the reliable performance of the top five predictors alone, this machine learning approach could help clinicians identify high-risk patients in clinical settings.

Optimal Landsat transforms for forest applications

NASA Technical Reports Server (NTRS)

Logan, T. L.; Strahler, A. H.

1983-01-01

Eleven transformations of data from four Landsat MSS channels were investigated to find if any of the transforms accentuated the separability of natural vegetation classes in regions of varying topographical relief. Attention was given to the divergence analysis and classification accuracy of information content of the eleven transforms and four channels. A useful scaling function was observed with the second eigenvector being the denominator in the divergence values obtained. The second eigenvector was found to reduce the effects of shadowing and differential illumination of vegetation signatures, thereby enhancing the divergence values. The highest accuracies in crop identification were provided by the averages of channels 4, 6, and 7 divided by the second eigenvector.

Land cover mapping in Latvia using hyperspectral airborne and simulated Sentinel-2 data

NASA Astrophysics Data System (ADS)

Jakovels, Dainis; Filipovs, Jevgenijs; Brauns, Agris; Taskovs, Juris; Erins, Gatis

2016-08-01

Land cover mapping in Latvia is performed as part of the Corine Land Cover (CLC) initiative every six years. The advantage of CLC is the creation of a standardized nomenclature and mapping protocol comparable across all European countries, thereby making it a valuable information source at the European level. However, low spatial resolution and accuracy, infrequent updates and expensive manual production has limited its use at the national level. As of now, there is no remote sensing based high resolution land cover and land use services designed specifically for Latvia which would account for the country's natural and land use specifics and end-user interests. The European Space Agency launched the Sentinel-2 satellite in 2015 aiming to provide continuity of free high resolution multispectral satellite data thereby presenting an opportunity to develop and adapted land cover and land use algorithm which accounts for national enduser needs. In this study, land cover mapping scheme according to national end-user needs was developed and tested in two pilot territories (Cesis and Burtnieki). Hyperspectral airborne data covering spectral range 400-2500 nm was acquired in summer 2015 using Airborne Surveillance and Environmental Monitoring System (ARSENAL). The gathered data was tested for land cover classification of seven general classes (urban/artificial, bare, forest, shrubland, agricultural/grassland, wetlands, water) and sub-classes specific for Latvia as well as simulation of Sentinel-2 satellite data. Hyperspectral data sets consist of 122 spectral bands in visible to near infrared spectral range (356-950 nm) and 100 bands in short wave infrared (950-2500 nm). Classification of land cover was tested separately for each sensor data and fused cross-sensor data. The best overall classification accuracy 84.2% and satisfactory classification accuracy (more than 80%) for 9 of 13 classes was obtained using Support Vector Machine (SVM) classifier with 109 band hyperspectral data. Grassland and agriculture land demonstrated lowest classification accuracy in pixel based approach, but result significantly improved by looking at agriculture polygons registered in Rural Support Service data as objects. The test of simulated Sentinel-2 bands for land cover mapping using SVM classifier showed 82.8% overall accuracy and satisfactory separation of 7 classes. SVM provided highest overall accuracy 84.2% in comparison to 75.9% for k-Nearest Neighbor and 79.2% Linear Discriminant Analysis classifiers.

Decoding motor responses from the EEG during altered states of consciousness induced by propofol

NASA Astrophysics Data System (ADS)

Blokland, Yvonne; Farquhar, Jason; Lerou, Jos; Mourisse, Jo; Scheffer, Gert Jan; van Geffen, Geert-Jan; Spyrou, Loukianos; Bruhn, Jörgen

2016-04-01

Objective. Patients undergoing general anesthesia may awaken and become aware of the surgical procedure. Due to neuromuscular blocking agents, patients could be conscious yet unable to move. Using brain-computer interface (BCI) technology, it may be possible to detect movement attempts from the EEG. However, it is unknown how an anesthetic influences the brain response to motor tasks. Approach. We tested the offline classification performance of a movement-based BCI in 12 healthy subjects at two effect-site concentrations of propofol. For each subject a second classifier was trained on the subject’s data obtained before sedation, then tested on the data obtained during sedation (‘transfer classification’). Main results. At concentration 0.5 μg ml-1, despite an overall propofol EEG effect, the mean single trial classification accuracy was 85% (95% CI 81%-89%), and 83% (79%-88%) for the transfer classification. At 1.0 μg ml-1, the accuracies were 81% (76%-86%), and 72% (66%-79%), respectively. At the highest propofol concentration for four subjects, unlike the remaining subjects, the movement-related brain response had been largely diminished, and the transfer classification accuracy was not significantly above chance. These subjects showed a slower and more erratic task response, indicating an altered state of consciousness distinct from that of the other subjects. Significance. The results show the potential of using a BCI to detect intra-operative awareness and justify further development of this paradigm. At the same time, the relationship between motor responses and consciousness and its clinical relevance for intraoperative awareness requires further investigation.

Performance comparison of machine learning algorithms and number of independent components used in fMRI decoding of belief vs. disbelief.

PubMed

Douglas, P K; Harris, Sam; Yuille, Alan; Cohen, Mark S

2011-05-15

Machine learning (ML) has become a popular tool for mining functional neuroimaging data, and there are now hopes of performing such analyses efficiently in real-time. Towards this goal, we compared accuracy of six different ML algorithms applied to neuroimaging data of persons engaged in a bivariate task, asserting their belief or disbelief of a variety of propositional statements. We performed unsupervised dimension reduction and automated feature extraction using independent component (IC) analysis and extracted IC time courses. Optimization of classification hyperparameters across each classifier occurred prior to assessment. Maximum accuracy was achieved at 92% for Random Forest, followed by 91% for AdaBoost, 89% for Naïve Bayes, 87% for a J48 decision tree, 86% for K*, and 84% for support vector machine. For real-time decoding applications, finding a parsimonious subset of diagnostic ICs might be useful. We used a forward search technique to sequentially add ranked ICs to the feature subspace. For the current data set, we determined that approximately six ICs represented a meaningful basis set for classification. We then projected these six IC spatial maps forward onto a later scanning session within subject. We then applied the optimized ML algorithms to these new data instances, and found that classification accuracy results were reproducible. Additionally, we compared our classification method to our previously published general linear model results on this same data set. The highest ranked IC spatial maps show similarity to brain regions associated with contrasts for belief > disbelief, and disbelief < belief. Copyright © 2010 Elsevier Inc. All rights reserved.

Comparing writing style feature-based classification methods for estimating user reputations in social media.

PubMed

Suh, Jong Hwan

2016-01-01

In recent years, the anonymous nature of the Internet has made it difficult to detect manipulated user reputations in social media, as well as to ensure the qualities of users and their posts. To deal with this, this study designs and examines an automatic approach that adopts writing style features to estimate user reputations in social media. Under varying ways of defining Good and Bad classes of user reputations based on the collected data, it evaluates the classification performance of the state-of-art methods: four writing style features, i.e. lexical, syntactic, structural, and content-specific, and eight classification techniques, i.e. four base learners-C4.5, Neural Network (NN), Support Vector Machine (SVM), and Naïve Bayes (NB)-and four Random Subspace (RS) ensemble methods based on the four base learners. When South Korea's Web forum, Daum Agora, was selected as a test bed, the experimental results show that the configuration of the full feature set containing content-specific features and RS-SVM combining RS and SVM gives the best accuracy for classification if the test bed poster reputations are segmented strictly into Good and Bad classes by portfolio approach. Pairwise t tests on accuracy confirm two expectations coming from the literature reviews: first, the feature set adding content-specific features outperform the others; second, ensemble learning methods are more viable than base learners. Moreover, among the four ways on defining the classes of user reputations, i.e. like, dislike, sum, and portfolio, the results show that the portfolio approach gives the highest accuracy.

Estimating local scaling properties for the classification of interstitial lung disease patterns

NASA Astrophysics Data System (ADS)

Huber, Markus B.; Nagarajan, Mahesh B.; Leinsinger, Gerda; Ray, Lawrence A.; Wismueller, Axel

2011-03-01

Local scaling properties of texture regions were compared in their ability to classify morphological patterns known as 'honeycombing' that are considered indicative for the presence of fibrotic interstitial lung diseases in high-resolution computed tomography (HRCT) images. For 14 patients with known occurrence of honeycombing, a stack of 70 axial, lung kernel reconstructed images were acquired from HRCT chest exams. 241 regions of interest of both healthy and pathological (89) lung tissue were identified by an experienced radiologist. Texture features were extracted using six properties calculated from gray-level co-occurrence matrices (GLCM), Minkowski Dimensions (MDs), and the estimation of local scaling properties with Scaling Index Method (SIM). A k-nearest-neighbor (k-NN) classifier and a Multilayer Radial Basis Functions Network (RBFN) were optimized in a 10-fold cross-validation for each texture vector, and the classification accuracy was calculated on independent test sets as a quantitative measure of automated tissue characterization. A Wilcoxon signed-rank test was used to compare two accuracy distributions including the Bonferroni correction. The best classification results were obtained by the set of SIM features, which performed significantly better than all the standard GLCM and MD features (p < 0.005) for both classifiers with the highest accuracy (94.1%, 93.7%; for the k-NN and RBFN classifier, respectively). The best standard texture features were the GLCM features 'homogeneity' (91.8%, 87.2%) and 'absolute value' (90.2%, 88.5%). The results indicate that advanced texture features using local scaling properties can provide superior classification performance in computer-assisted diagnosis of interstitial lung diseases when compared to standard texture analysis methods.

Feature Selection for Classification of Polar Regions Using a Fuzzy Expert System

NASA Technical Reports Server (NTRS)

Penaloza, Mauel A.; Welch, Ronald M.

1996-01-01

Labeling, feature selection, and the choice of classifier are critical elements for classification of scenes and for image understanding. This study examines several methods for feature selection in polar regions, including the list, of a fuzzy logic-based expert system for further refinement of a set of selected features. Six Advanced Very High Resolution Radiometer (AVHRR) Local Area Coverage (LAC) arctic scenes are classified into nine classes: water, snow / ice, ice cloud, land, thin stratus, stratus over water, cumulus over water, textured snow over water, and snow-covered mountains. Sixty-seven spectral and textural features are computed and analyzed by the feature selection algorithms. The divergence, histogram analysis, and discriminant analysis approaches are intercompared for their effectiveness in feature selection. The fuzzy expert system method is used not only to determine the effectiveness of each approach in classifying polar scenes, but also to further reduce the features into a more optimal set. For each selection method,features are ranked from best to worst, and the best half of the features are selected. Then, rules using these selected features are defined. The results of running the fuzzy expert system with these rules show that the divergence method produces the best set features, not only does it produce the highest classification accuracy, but also it has the lowest computation requirements. A reduction of the set of features produced by the divergence method using the fuzzy expert system results in an overall classification accuracy of over 95 %. However, this increase of accuracy has a high computation cost.

[Extracting black soil border in Heilongjiang province based on spectral angle match method].

PubMed

Zhang, Xin-Le; Zhang, Shu-Wen; Li, Ying; Liu, Huan-Jun

2009-04-01

As soils are generally covered by vegetation most time of a year, the spectral reflectance collected by remote sensing technique is from the mixture of soil and vegetation, so the classification precision based on remote sensing (RS) technique is unsatisfied. Under RS and geographic information systems (GIS) environment and with the help of buffer and overlay analysis methods, land use and soil maps were used to derive regions of interest (ROI) for RS supervised classification, which plus MODIS reflectance products were chosen to extract black soil border, with methods including spectral single match. The results showed that the black soil border in Heilongjiang province can be extracted with soil remote sensing method based on MODIS reflectance products, especially in the north part of black soil zone; the classification precision of spectral angel mapping method is the highest, but the classifying accuracy of other soils can not meet the need, because of vegetation covering and similar spectral characteristics; even for the same soil, black soil, the classifying accuracy has obvious spatial heterogeneity, in the north part of black soil zone in Heilongjiang province it is higher than in the south, which is because of spectral differences; as soil uncovering period in Northeastern China is relatively longer, high temporal resolution make MODIS images get the advantage over soil remote sensing classification; with the help of GIS, extracting ROIs by making the best of auxiliary data can improve the precision of soil classification; with the help of auxiliary information, such as topography and climate, the classification accuracy was enhanced significantly. As there are five main factors determining soil classes, much data of different types, such as DEM, terrain factors, climate (temperature, precipitation, etc.), parent material, vegetation map, and remote sensing images, were introduced to classify soils, so how to choose some of the data and quantify the weights of different data layers needs further study.

Design of Embedded System for Multivariate Classification of Finger and Thumb Movements Using EEG Signals for Control of Upper Limb Prosthesis

PubMed Central

Javed, Amna; Tiwana, Mohsin I.; Khan, Umar Shahbaz

2018-01-01

Brain Computer Interface (BCI) determines the intent of the user from a variety of electrophysiological signals. These signals, Slow Cortical Potentials, are recorded from scalp, and cortical neuronal activity is recorded by implanted electrodes. This paper is focused on design of an embedded system that is used to control the finger movements of an upper limb prosthesis using Electroencephalogram (EEG) signals. This is a follow-up of our previous research which explored the best method to classify three movements of fingers (thumb movement, index finger movement, and first movement). Two-stage logistic regression classifier exhibited the highest classification accuracy while Power Spectral Density (PSD) was used as a feature of the filtered signal. The EEG signal data set was recorded using a 14-channel electrode headset (a noninvasive BCI system) from right-handed, neurologically intact volunteers. Mu (commonly known as alpha waves) and Beta Rhythms (8–30 Hz) containing most of the movement data were retained through filtering using “Arduino Uno” microcontroller followed by 2-stage logistic regression to obtain a mean classification accuracy of 70%. PMID:29888252

Texture analysis of common renal masses in multiple MR sequences for prediction of pathology

NASA Astrophysics Data System (ADS)

Hoang, Uyen N.; Malayeri, Ashkan A.; Lay, Nathan S.; Summers, Ronald M.; Yao, Jianhua

2017-03-01

This pilot study performs texture analysis on multiple magnetic resonance (MR) images of common renal masses for differentiation of renal cell carcinoma (RCC). Bounding boxes are drawn around each mass on one axial slice in T1 delayed sequence to use for feature extraction and classification. All sequences (T1 delayed, venous, arterial, pre-contrast phases, T2, and T2 fat saturated sequences) are co-registered and texture features are extracted from each sequence simultaneously. Random forest is used to construct models to classify lesions on 96 normal regions, 87 clear cell RCCs, 8 papillary RCCs, and 21 renal oncocytomas; ground truths are verified through pathology reports. The highest performance is seen in random forest model when data from all sequences are used in conjunction, achieving an overall classification accuracy of 83.7%. When using data from one single sequence, the overall accuracies achieved for T1 delayed, venous, arterial, and pre-contrast phase, T2, and T2 fat saturated were 79.1%, 70.5%, 56.2%, 61.0%, 60.0%, and 44.8%, respectively. This demonstrates promising results of utilizing intensity information from multiple MR sequences for accurate classification of renal masses.

High-Throughput Classification of Radiographs Using Deep Convolutional Neural Networks.

PubMed

Rajkomar, Alvin; Lingam, Sneha; Taylor, Andrew G; Blum, Michael; Mongan, John

2017-02-01

The study aimed to determine if computer vision techniques rooted in deep learning can use a small set of radiographs to perform clinically relevant image classification with high fidelity. One thousand eight hundred eighty-five chest radiographs on 909 patients obtained between January 2013 and July 2015 at our institution were retrieved and anonymized. The source images were manually annotated as frontal or lateral and randomly divided into training, validation, and test sets. Training and validation sets were augmented to over 150,000 images using standard image manipulations. We then pre-trained a series of deep convolutional networks based on the open-source GoogLeNet with various transformations of the open-source ImageNet (non-radiology) images. These trained networks were then fine-tuned using the original and augmented radiology images. The model with highest validation accuracy was applied to our institutional test set and a publicly available set. Accuracy was assessed by using the Youden Index to set a binary cutoff for frontal or lateral classification. This retrospective study was IRB approved prior to initiation. A network pre-trained on 1.2 million greyscale ImageNet images and fine-tuned on augmented radiographs was chosen. The binary classification method correctly classified 100 % (95 % CI 99.73-100 %) of both our test set and the publicly available images. Classification was rapid, at 38 images per second. A deep convolutional neural network created using non-radiological images, and an augmented set of radiographs is effective in highly accurate classification of chest radiograph view type and is a feasible, rapid method for high-throughput annotation.

A Machine Learning-based Method for Question Type Classification in Biomedical Question Answering.

PubMed

Sarrouti, Mourad; Ouatik El Alaoui, Said

2017-05-18

Biomedical question type classification is one of the important components of an automatic biomedical question answering system. The performance of the latter depends directly on the performance of its biomedical question type classification system, which consists of assigning a category to each question in order to determine the appropriate answer extraction algorithm. This study aims to automatically classify biomedical questions into one of the four categories: (1) yes/no, (2) factoid, (3) list, and (4) summary. In this paper, we propose a biomedical question type classification method based on machine learning approaches to automatically assign a category to a biomedical question. First, we extract features from biomedical questions using the proposed handcrafted lexico-syntactic patterns. Then, we feed these features for machine-learning algorithms. Finally, the class label is predicted using the trained classifiers. Experimental evaluations performed on large standard annotated datasets of biomedical questions, provided by the BioASQ challenge, demonstrated that our method exhibits significant improved performance when compared to four baseline systems. The proposed method achieves a roughly 10-point increase over the best baseline in terms of accuracy. Moreover, the obtained results show that using handcrafted lexico-syntactic patterns as features' provider of support vector machine (SVM) lead to the highest accuracy of 89.40 %. The proposed method can automatically classify BioASQ questions into one of the four categories: yes/no, factoid, list, and summary. Furthermore, the results demonstrated that our method produced the best classification performance compared to four baseline systems.

Multimodal 2D Brain Computer Interface.

PubMed

Almajidy, Rand K; Boudria, Yacine; Hofmann, Ulrich G; Besio, Walter; Mankodiya, Kunal

2015-08-01

In this work we used multimodal, non-invasive brain signal recording systems, namely Near Infrared Spectroscopy (NIRS), disc electrode electroencephalography (EEG) and tripolar concentric ring electrodes (TCRE) electroencephalography (tEEG). 7 healthy subjects participated in our experiments to control a 2-D Brain Computer Interface (BCI). Four motor imagery task were performed, imagery motion of the left hand, the right hand, both hands and both feet. The signal slope (SS) of the change in oxygenated hemoglobin concentration measured by NIRS was used for feature extraction while the power spectrum density (PSD) of both EEG and tEEG in the frequency band 8-30Hz was used for feature extraction. Linear Discriminant Analysis (LDA) was used to classify different combinations of the aforementioned features. The highest classification accuracy (85.2%) was achieved by using features from all the three brain signals recording modules. The improvement in classification accuracy was highly significant (p = 0.0033) when using the multimodal signals features as compared to pure EEG features.

A combined LS-SVM & MLR QSAR workflow for predicting the inhibition of CXCR3 receptor by quinazolinone analogs.

PubMed

Afantitis, Antreas; Melagraki, Georgia; Sarimveis, Haralambos; Koutentis, Panayiotis A; Igglessi-Markopoulou, Olga; Kollias, George

2010-05-01

A novel QSAR workflow is constructed that combines MLR with LS-SVM classification techniques for the identification of quinazolinone analogs as "active" or "non-active" CXCR3 antagonists. The accuracy of the LS-SVM classification technique for the training set and test was 100% and 90%, respectively. For the "active" analogs a validated MLR QSAR model estimates accurately their I-IP10 IC(50) inhibition values. The accuracy of the QSAR model (R (2) = 0.80) is illustrated using various evaluation techniques, such as leave-one-out procedure (R(LOO2)) = 0.67) and validation through an external test set (R(pred2) = 0.78). The key conclusion of this study is that the selected molecular descriptors, Highest Occupied Molecular Orbital energy (HOMO), Principal Moment of Inertia along X and Y axes PMIX and PMIZ, Polar Surface Area (PSA), Presence of triple bond (PTrplBnd), and Kier shape descriptor ((1) kappa), demonstrate discriminatory and pharmacophore abilities.

Automated sleep scoring and sleep apnea detection in children

NASA Astrophysics Data System (ADS)

Baraglia, David P.; Berryman, Matthew J.; Coussens, Scott W.; Pamula, Yvonne; Kennedy, Declan; Martin, A. James; Abbott, Derek

2005-12-01

This paper investigates the automated detection of a patient's breathing rate and heart rate from their skin conductivity as well as sleep stage scoring and breathing event detection from their EEG. The software developed for these tasks is tested on data sets obtained from the sleep disorders unit at the Adelaide Women's and Children's Hospital. The sleep scoring and breathing event detection tasks used neural networks to achieve signal classification. The Fourier transform and the Higuchi fractal dimension were used to extract features for input to the neural network. The filtered skin conductivity appeared visually to bear a similarity to the breathing and heart rate signal, but a more detailed evaluation showed the relation was not consistent. Sleep stage classification was achieved with and accuracy of around 65% with some stages being accurately scored and others poorly scored. The two breathing events hypopnea and apnea were scored with varying degrees of accuracy with the highest scores being around 75% and 30%.

Automatic classification of 6-month-old infants at familial risk for language-based learning disorder using a support vector machine.

PubMed

Zare, Marzieh; Rezvani, Zahra; Benasich, April A

2016-07-01

This study assesses the ability of a novel, "automatic classification" approach to facilitate identification of infants at highest familial risk for language-learning disorders (LLD) and to provide converging assessments to enable earlier detection of developmental disorders that disrupt language acquisition. Network connectivity measures derived from 62-channel electroencephalogram (EEG) recording were used to identify selected features within two infant groups who differed on LLD risk: infants with a family history of LLD (FH+) and typically-developing infants without such a history (FH-). A support vector machine was deployed; global efficiency and global and local clustering coefficients were computed. A novel minimum spanning tree (MST) approach was also applied. Cross-validation was employed to assess the resultant classification. Infants were classified with about 80% accuracy into FH+ and FH- groups with 89% specificity and precision of 92%. Clustering patterns differed by risk group and MST network analysis suggests that FH+ infants' EEG complexity patterns were significantly different from FH- infants. The automatic classification techniques used here were shown to be both robust and reliable and should provide valuable information when applied to early identification of risk or clinical groups. The ability to identify infants at highest risk for LLD using "automatic classification" strategies is a novel convergent approach that may facilitate earlier diagnosis and remediation. Copyright © 2016 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.

Qualitative properties of roasting defect beans and development of its classification methods by hyperspectral imaging technology.

PubMed

Cho, Jeong-Seok; Bae, Hyung-Jin; Cho, Byoung-Kwan; Moon, Kwang-Deog

2017-04-01

Qualitative properties of roasting defect coffee beans and their classification methods were studied using hyperspectral imaging (HSI). The roasting defect beans were divided into 5 groups: medium roasting (Cont), under developed (RD-1), over roasting (RD-2), interior under developed (RD-3), and interior scorching (RD-4). The following qualitative properties were assayed: browning index (BI), moisture content (MC), chlorogenic acid (CA), trigonelline (TG), and caffeine (CF) content. Their HSI spectra (1000-1700nm) were also analysed to develop the classification methods of roasting defect beans. RD-2 showed the highest BI and the lowest MC, CA, and TG content. The accuracy of classification model of partial least-squares discriminant was 86.2%. The most powerful wavelength to classify the defective beans was approximately 1420nm (related to OH bond). The HSI reflectance values at 1420nm showed similar tendency with MC, enabling the use of this technology to classify the roasting defect beans. Copyright © 2016. Published by Elsevier Ltd.

Driver Fatigue Classification With Independent Component by Entropy Rate Bound Minimization Analysis in an EEG-Based System.

PubMed

Chai, Rifai; Naik, Ganesh R; Nguyen, Tuan Nghia; Ling, Sai Ho; Tran, Yvonne; Craig, Ashley; Nguyen, Hung T

2017-05-01

This paper presents a two-class electroencephal-ography-based classification for classifying of driver fatigue (fatigue state versus alert state) from 43 healthy participants. The system uses independent component by entropy rate bound minimization analysis (ERBM-ICA) for the source separation, autoregressive (AR) modeling for the features extraction, and Bayesian neural network for the classification algorithm. The classification results demonstrate a sensitivity of 89.7%, a specificity of 86.8%, and an accuracy of 88.2%. The combination of ERBM-ICA (source separator), AR (feature extractor), and Bayesian neural network (classifier) provides the best outcome with a p-value < 0.05 with the highest value of area under the receiver operating curve (AUC-ROC = 0.93) against other methods such as power spectral density as feature extractor (AUC-ROC = 0.81). The results of this study suggest the method could be utilized effectively for a countermeasure device for driver fatigue identification and other adverse event applications.

The Cross-Entropy Based Multi-Filter Ensemble Method for Gene Selection.

PubMed

Sun, Yingqiang; Lu, Chengbo; Li, Xiaobo

2018-05-17

The gene expression profile has the characteristics of a high dimension, low sample, and continuous type, and it is a great challenge to use gene expression profile data for the classification of tumor samples. This paper proposes a cross-entropy based multi-filter ensemble (CEMFE) method for microarray data classification. Firstly, multiple filters are used to select the microarray data in order to obtain a plurality of the pre-selected feature subsets with a different classification ability. The top N genes with the highest rank of each subset are integrated so as to form a new data set. Secondly, the cross-entropy algorithm is used to remove the redundant data in the data set. Finally, the wrapper method, which is based on forward feature selection, is used to select the best feature subset. The experimental results show that the proposed method is more efficient than other gene selection methods and that it can achieve a higher classification accuracy under fewer characteristic genes.

Temporal optimisation of image acquisition for land cover classification with Random Forest and MODIS time-series

NASA Astrophysics Data System (ADS)

Nitze, Ingmar; Barrett, Brian; Cawkwell, Fiona

2015-02-01

The analysis and classification of land cover is one of the principal applications in terrestrial remote sensing. Due to the seasonal variability of different vegetation types and land surface characteristics, the ability to discriminate land cover types changes over time. Multi-temporal classification can help to improve the classification accuracies, but different constraints, such as financial restrictions or atmospheric conditions, may impede their application. The optimisation of image acquisition timing and frequencies can help to increase the effectiveness of the classification process. For this purpose, the Feature Importance (FI) measure of the state-of-the art machine learning method Random Forest was used to determine the optimal image acquisition periods for a general (Grassland, Forest, Water, Settlement, Peatland) and Grassland specific (Improved Grassland, Semi-Improved Grassland) land cover classification in central Ireland based on a 9-year time-series of MODIS Terra 16 day composite data (MOD13Q1). Feature Importances for each acquisition period of the Enhanced Vegetation Index (EVI) and Normalised Difference Vegetation Index (NDVI) were calculated for both classification scenarios. In the general land cover classification, the months December and January showed the highest, and July and August the lowest separability for both VIs over the entire nine-year period. This temporal separability was reflected in the classification accuracies, where the optimal choice of image dates outperformed the worst image date by 13% using NDVI and 5% using EVI on a mono-temporal analysis. With the addition of the next best image periods to the data input the classification accuracies converged quickly to their limit at around 8-10 images. The binary classification schemes, using two classes only, showed a stronger seasonal dependency with a higher intra-annual, but lower inter-annual variation. Nonetheless anomalous weather conditions, such as the cold winter of 2009/2010 can alter the temporal separability pattern significantly. Due to the extensive use of the NDVI for land cover discrimination, the findings of this study should be transferrable to data from other optical sensors with a higher spatial resolution. However, the high impact of outliers from the general climatic pattern highlights the limitation of spatial transferability to locations with different climatic and land cover conditions. The use of high-temporal, moderate resolution data such as MODIS in conjunction with machine-learning techniques proved to be a good base for the prediction of image acquisition timing for optimal land cover classification results.

Optimal classification for the diagnosis of duchenne muscular dystrophy images using support vector machines.

PubMed

Zhang, Ming-Huan; Ma, Jun-Shan; Shen, Ying; Chen, Ying

2016-09-01

This study aimed to investigate the optimal support vector machines (SVM)-based classifier of duchenne muscular dystrophy (DMD) magnetic resonance imaging (MRI) images. T1-weighted (T1W) and T2-weighted (T2W) images of the 15 boys with DMD and 15 normal controls were obtained. Textural features of the images were extracted and wavelet decomposed, and then, principal features were selected. Scale transform was then performed for MRI images. Afterward, SVM-based classifiers of MRI images were analyzed based on the radical basis function and decomposition levels. The cost (C) parameter and kernel parameter [Formula: see text] were used for classification. Then, the optimal SVM-based classifier, expressed as [Formula: see text]), was identified by performance evaluation (sensitivity, specificity and accuracy). Eight of 12 textural features were selected as principal features (eigenvalues [Formula: see text]). The 16 SVM-based classifiers were obtained using combination of (C, [Formula: see text]), and those with lower C and [Formula: see text] values showed higher performances, especially classifier of [Formula: see text]). The SVM-based classifiers of T1W images showed higher performance than T1W images at the same decomposition level. The T1W images in classifier of [Formula: see text]) at level 2 decomposition showed the highest performance of all, and its overall correct sensitivity, specificity, and accuracy reached 96.9, 97.3, and 97.1 %, respectively. The T1W images in SVM-based classifier [Formula: see text] at level 2 decomposition showed the highest performance of all, demonstrating that it was the optimal classification for the diagnosis of DMD.

Variance approximations for assessments of classification accuracy

Treesearch

R. L. Czaplewski

1994-01-01

Variance approximations are derived for the weighted and unweighted kappa statistics, the conditional kappa statistic, and conditional probabilities. These statistics are useful to assess classification accuracy, such as accuracy of remotely sensed classifications in thematic maps when compared to a sample of reference classifications made in the field. Published...

Classification of interstitial lung disease patterns with topological texture features

NASA Astrophysics Data System (ADS)

Huber, Markus B.; Nagarajan, Mahesh; Leinsinger, Gerda; Ray, Lawrence A.; Wismüller, Axel

2010-03-01

Topological texture features were compared in their ability to classify morphological patterns known as 'honeycombing' that are considered indicative for the presence of fibrotic interstitial lung diseases in high-resolution computed tomography (HRCT) images. For 14 patients with known occurrence of honey-combing, a stack of 70 axial, lung kernel reconstructed images were acquired from HRCT chest exams. A set of 241 regions of interest of both healthy and pathological (89) lung tissue were identified by an experienced radiologist. Texture features were extracted using six properties calculated from gray-level co-occurrence matrices (GLCM), Minkowski Dimensions (MDs), and three Minkowski Functionals (MFs, e.g. MF.euler). A k-nearest-neighbor (k-NN) classifier and a Multilayer Radial Basis Functions Network (RBFN) were optimized in a 10-fold cross-validation for each texture vector, and the classification accuracy was calculated on independent test sets as a quantitative measure of automated tissue characterization. A Wilcoxon signed-rank test was used to compare two accuracy distributions and the significance thresholds were adjusted for multiple comparisons by the Bonferroni correction. The best classification results were obtained by the MF features, which performed significantly better than all the standard GLCM and MD features (p < 0.005) for both classifiers. The highest accuracy was found for MF.euler (97.5%, 96.6%; for the k-NN and RBFN classifier, respectively). The best standard texture features were the GLCM features 'homogeneity' (91.8%, 87.2%) and 'absolute value' (90.2%, 88.5%). The results indicate that advanced topological texture features can provide superior classification performance in computer-assisted diagnosis of interstitial lung diseases when compared to standard texture analysis methods.

Comparison of dual-biomarker PIB-PET and dual-tracer PET in AD diagnosis.

PubMed

Fu, Liping; Liu, Linwen; Zhang, Jinming; Xu, Baixuan; Fan, Yong; Tian, Jiahe

2014-11-01

To identify the optimal time window for capturing perfusion information from early (11)C-PIB imaging frames (perfusion PIB, (11)C-pPIB) and to compare the performance of (18)F-FDG PET and "dual biomarker" (11)C-PIB PET [(11)C-pPIB and amyloid PIB ((11)C-aPIB)] for classification of AD, MCI and CN subjects. Forty subjects (14 CN, 12 MCI and 14 AD patients) underwent (18)F-FDG and (11)C-PIB PET studies. Pearson correlation between the (18)F-FDG image and sum of early (11)C-PIB frames was maximised to identify the optimal time window for (11)C-pPIB. The classification power of imaging parameters was evaluated with a leave-one-out validation. A 7-min time window yielded the highest correlation between (18)F-FDG and (11)C-pPIB. (11)C-pPIB and (18)F-FDG images shared a similar radioactive distribution pattern. (18)F-FDG performed better than (11)C-pPIB for the classification of both AD vs. CN and MCI vs. CN. (11)C-pPIB + (11)C-aPIB and (18)F-FDG + (11)C-aPIB yielded the highest classification accuracy for the classification of AD vs. CN, and (18)F-FDG + (11)C-aPIB had the best classification performance for the classification of MCI vs. C-pPIB could serve as a useful biomarker of rCBF for measuring neural activity and improve the diagnostic power of PET for AD in conjunction with (11)C-aPIB. (18)F-FDG and (11)C-PIB dual-tracer PET examination could better detect MCI. • Dual-tracer PET examination provides neurofunctional and neuropathological information for AD diagnosis. • The identified optimal 11C-pPIB time frames had highest correlation with 18F-FDG. • 11C-pPIB images shared a similar radioactive distribution pattern with 18F-FDG images. • 11C-pPIB can provide neurofunctional information. • Dual-tracer PET examination could better detect MCI.

Serial combination of non-invasive tools improves the diagnostic accuracy of severe liver fibrosis in patients with NAFLD.

PubMed

Petta, S; Wong, V W-S; Cammà, C; Hiriart, J-B; Wong, G L-H; Vergniol, J; Chan, A W-H; Di Marco, V; Merrouche, W; Chan, H L-Y; Marra, F; Le-Bail, B; Arena, U; Craxì, A; de Ledinghen, V

2017-09-01

The accuracy of available non-invasive tools for staging severe fibrosis in patients with nonalcoholic fatty liver disease (NAFLD) is still limited. To assess the diagnostic performance of paired or serial combination of non-invasive tools in NAFLD patients. We analysed data from 741 patients with a histological diagnosis of NAFLD. The GGT/PLT, APRI, AST/ALT, BARD, FIB-4, and NAFLD Fibrosis Score (NFS) scores were calculated according to published algorithms. Liver stiffness measurement (LSM) was performed by FibroScan. LSM, NFS and FIB-4 were the best non-invasive tools for staging F3-F4 fibrosis (AUC 0.863, 0.774, and 0.792, respectively), with LSM having the highest sensitivity (90%), and the highest NPV (94%), and NFS and FIB-4 the highest specificity (97% and 93%, respectively), and the highest PPV (73% and 79%, respectively). The paired combination of LSM or NFS with FIB-4 strongly reduced the likelihood of wrongly classified patients (ranging from 2.7% to 2.6%), at the price of a high uncertainty area (ranging from 54.1% to 58.2%), and of a low overall accuracy (ranging from 43% to 39.1%). The serial combination with the second test used in patients in the grey area of the first test and in those with high LSM values (>9.6 KPa) or low NFS or FIB-4 values (<-1.455 and <1.30, respectively) overall increased the diagnostic performance generating an accuracy ranging from 69.8% to 70.1%, an uncertainty area ranging from 18.9% to 20.4% and a rate of wrong classification ranging from 9.2% to 11.3%. The serial combination of LSM with FIB-4/NFS has a good diagnostic accuracy for the non-invasive diagnosis of severe fibrosis in NAFLD. © 2017 John Wiley & Sons Ltd.

Classifying coastal resources by integrating optical and radar imagery and color infrared photography

USGS Publications Warehouse

Ramsey, Elijah W.; Nelson, Gene A.; Sapkota, Sijan

1998-01-01

A progressive classification of a marsh and forest system using Landsat Thematic Mapper (TM), color infrared (CIR) photograph, and ERS-1 synthetic aperture radar (SAR) data improved classification accuracy when compared to classification using solely TM reflective band data. The classification resulted in a detailed identification of differences within a nearly monotypic black needlerush marsh. Accuracy percentages of these classes were surprisingly high given the complexities of classification. The detailed classification resulted in a more accurate portrayal of the marsh transgressive sequence than was obtainable with TM data alone. Individual sensor contribution to the improved classification was compared to that using only the six reflective TM bands. Individually, the green reflective CIR and SAR data identified broad categories of water, marsh, and forest. In combination with TM, SAR and the green CIR band each improved overall accuracy by about 3% and 15% respectively. The SAR data improved the TM classification accuracy mostly in the marsh classes. The green CIR data also improved the marsh classification accuracy and accuracies in some water classes. The final combination of all sensor data improved almost all class accuracies from 2% to 70% with an overall improvement of about 20% over TM data alone. Not only was the identification of vegetation types improved, but the spatial detail of the classification approached 10 m in some areas.

Comparison of Feature Selection Techniques in Machine Learning for Anatomical Brain MRI in Dementia.

PubMed

Tohka, Jussi; Moradi, Elaheh; Huttunen, Heikki

2016-07-01

We present a comparative split-half resampling analysis of various data driven feature selection and classification methods for the whole brain voxel-based classification analysis of anatomical magnetic resonance images. We compared support vector machines (SVMs), with or without filter based feature selection, several embedded feature selection methods and stability selection. While comparisons of the accuracy of various classification methods have been reported previously, the variability of the out-of-training sample classification accuracy and the set of selected features due to independent training and test sets have not been previously addressed in a brain imaging context. We studied two classification problems: 1) Alzheimer's disease (AD) vs. normal control (NC) and 2) mild cognitive impairment (MCI) vs. NC classification. In AD vs. NC classification, the variability in the test accuracy due to the subject sample did not vary between different methods and exceeded the variability due to different classifiers. In MCI vs. NC classification, particularly with a large training set, embedded feature selection methods outperformed SVM-based ones with the difference in the test accuracy exceeding the test accuracy variability due to the subject sample. The filter and embedded methods produced divergent feature patterns for MCI vs. NC classification that suggests the utility of the embedded feature selection for this problem when linked with the good generalization performance. The stability of the feature sets was strongly correlated with the number of features selected, weakly correlated with the stability of classification accuracy, and uncorrelated with the average classification accuracy.

Attribute-Level and Pattern-Level Classification Consistency and Accuracy Indices for Cognitive Diagnostic Assessment

ERIC Educational Resources Information Center

Wang, Wenyi; Song, Lihong; Chen, Ping; Meng, Yaru; Ding, Shuliang

2015-01-01

Classification consistency and accuracy are viewed as important indicators for evaluating the reliability and validity of classification results in cognitive diagnostic assessment (CDA). Pattern-level classification consistency and accuracy indices were introduced by Cui, Gierl, and Chang. However, the indices at the attribute level have not yet…

Classification of suicide attempters in schizophrenia using sociocultural and clinical features: A machine learning approach.

PubMed

Hettige, Nuwan C; Nguyen, Thai Binh; Yuan, Chen; Rajakulendran, Thanara; Baddour, Jermeen; Bhagwat, Nikhil; Bani-Fatemi, Ali; Voineskos, Aristotle N; Mallar Chakravarty, M; De Luca, Vincenzo

2017-07-01

Suicide is a major concern for those afflicted by schizophrenia. Identifying patients at the highest risk for future suicide attempts remains a complex problem for psychiatric interventions. Machine learning models allow for the integration of many risk factors in order to build an algorithm that predicts which patients are likely to attempt suicide. Currently it is unclear how to integrate previously identified risk factors into a clinically relevant predictive tool to estimate the probability of a patient with schizophrenia for attempting suicide. We conducted a cross-sectional assessment on a sample of 345 participants diagnosed with schizophrenia spectrum disorders. Suicide attempters and non-attempters were clearly identified using the Columbia Suicide Severity Rating Scale (C-SSRS) and the Beck Suicide Ideation Scale (BSS). We developed four classification algorithms using a regularized regression, random forest, elastic net and support vector machine models with sociocultural and clinical variables as features to train the models. All classification models performed similarly in identifying suicide attempters and non-attempters. Our regularized logistic regression model demonstrated an accuracy of 67% and an area under the curve (AUC) of 0.71, while the random forest model demonstrated 66% accuracy and an AUC of 0.67. Support vector classifier (SVC) model demonstrated an accuracy of 67% and an AUC of 0.70, and the elastic net model demonstrated and accuracy of 65% and an AUC of 0.71. Machine learning algorithms offer a relatively successful method for incorporating many clinical features to predict individuals at risk for future suicide attempts. Increased performance of these models using clinically relevant variables offers the potential to facilitate early treatment and intervention to prevent future suicide attempts. Copyright © 2017 Elsevier Inc. All rights reserved.

A hybrid clustering and classification approach for predicting crash injury severity on rural roads.

PubMed

Hasheminejad, Seyed Hessam-Allah; Zahedi, Mohsen; Hasheminejad, Seyed Mohammad Hossein

2018-03-01

As a threat for transportation system, traffic crashes have a wide range of social consequences for governments. Traffic crashes are increasing in developing countries and Iran as a developing country is not immune from this risk. There are several researches in the literature to predict traffic crash severity based on artificial neural networks (ANNs), support vector machines and decision trees. This paper attempts to investigate the crash injury severity of rural roads by using a hybrid clustering and classification approach to compare the performance of classification algorithms before and after applying the clustering. In this paper, a novel rule-based genetic algorithm (GA) is proposed to predict crash injury severity, which is evaluated by performance criteria in comparison with classification algorithms like ANN. The results obtained from analysis of 13,673 crashes (5600 property damage, 778 fatal crashes, 4690 slight injuries and 2605 severe injuries) on rural roads in Tehran Province of Iran during 2011-2013 revealed that the proposed GA method outperforms other classification algorithms based on classification metrics like precision (86%), recall (88%) and accuracy (87%). Moreover, the proposed GA method has the highest level of interpretation, is easy to understand and provides feedback to analysts.

Wechsler Intelligence Scale for Children-fourth edition (WISC-IV) short-form validity: a comparison study in pediatric epilepsy.

PubMed

Hrabok, Marianne; Brooks, Brian L; Fay-McClymont, Taryn B; Sherman, Elisabeth M S

2014-01-01

The purpose of this article was to investigate the accuracy of the WISC-IV short forms in estimating Full Scale Intelligence Quotient (FSIQ) and General Ability Index (GAI) in pediatric epilepsy. One hundred and four children with epilepsy completed the WISC-IV as part of a neuropsychological assessment at a tertiary-level children's hospital. The clinical accuracy of eight short forms was assessed in two ways: (a) accuracy within +/- 5 index points of FSIQ and (b) the clinical classification rate according to Wechsler conventions. The sample was further subdivided into low FSIQ (≤ 80) and high FSIQ (> 80). All short forms were significantly correlated with FSIQ. Seven-subtest (Crawford et al. [2010] FSIQ) and 5-subtest (BdSiCdVcLn) short forms yielded the highest clinical accuracy rates (77%-89%). Overall, a 2-subtest (VcMr) short form yielded the lowest clinical classification rates for FSIQ (35%-63%). The short form yielding the most accurate estimate of GAI was VcSiMrBd (73%-84%). Short forms show promise as useful estimates. The 7-subtest (Crawford et al., 2010) and 5-subtest (BdSiVcLnCd) short forms yielded the most accurate estimates of FSIQ. VcSiMrBd yielded the most accurate estimate of GAI. Clinical recommendations are provided for use of short forms in pediatric epilepsy.

The Accuracy and Reliability of Crowdsource Annotations of Digital Retinal Images

PubMed Central

Mitry, Danny; Zutis, Kris; Dhillon, Baljean; Peto, Tunde; Hayat, Shabina; Khaw, Kay-Tee; Morgan, James E.; Moncur, Wendy; Trucco, Emanuele; Foster, Paul J.

2016-01-01

Purpose Crowdsourcing is based on outsourcing computationally intensive tasks to numerous individuals in the online community who have no formal training. Our aim was to develop a novel online tool designed to facilitate large-scale annotation of digital retinal images, and to assess the accuracy of crowdsource grading using this tool, comparing it to expert classification. Methods We used 100 retinal fundus photograph images with predetermined disease criteria selected by two experts from a large cohort study. The Amazon Mechanical Turk Web platform was used to drive traffic to our site so anonymous workers could perform a classification and annotation task of the fundus photographs in our dataset after a short training exercise. Three groups were assessed: masters only, nonmasters only and nonmasters with compulsory training. We calculated the sensitivity, specificity, and area under the curve (AUC) of receiver operating characteristic (ROC) plots for all classifications compared to expert grading, and used the Dice coefficient and consensus threshold to assess annotation accuracy. Results In total, we received 5389 annotations for 84 images (excluding 16 training images) in 2 weeks. A specificity and sensitivity of 71% (95% confidence interval [CI], 69%–74%) and 87% (95% CI, 86%–88%) was achieved for all classifications. The AUC in this study for all classifications combined was 0.93 (95% CI, 0.91–0.96). For image annotation, a maximal Dice coefficient (∼0.6) was achieved with a consensus threshold of 0.25. Conclusions This study supports the hypothesis that annotation of abnormalities in retinal images by ophthalmologically naive individuals is comparable to expert annotation. The highest AUC and agreement with expert annotation was achieved in the nonmasters with compulsory training group. Translational Relevance The use of crowdsourcing as a technique for retinal image analysis may be comparable to expert graders and has the potential to deliver timely, accurate, and cost-effective image analysis. PMID:27668130

Multi-phenology WorldView-2 imagery improves remote sensing of savannah tree species

NASA Astrophysics Data System (ADS)

Madonsela, Sabelo; Cho, Moses Azong; Mathieu, Renaud; Mutanga, Onisimo; Ramoelo, Abel; Kaszta, Żaneta; Kerchove, Ruben Van De; Wolff, Eléonore

2017-06-01

Biodiversity mapping in African savannah is important for monitoring changes and ensuring sustainable use of ecosystem resources. Biodiversity mapping can benefit from multi-spectral instruments such as WorldView-2 with very high spatial resolution and a spectral configuration encompassing important spectral regions not previously available for vegetation mapping. This study investigated i) the benefits of the eight-band WorldView-2 (WV-2) spectral configuration for discriminating tree species in Southern African savannah and ii) if multiple-images acquired at key points of the typical phenological development of savannahs (peak productivity, transition to senescence) improve on tree species classifications. We first assessed the discriminatory power of WV-2 bands using interspecies-Spectral Angle Mapper (SAM) via Band Add-On procedure and tested the spectral capability of WorldView-2 against simulated IKONOS for tree species classification. The results from interspecies-SAM procedure identified the yellow and red bands as the most statistically significant bands (p = 0.000251 and p = 0.000039 respectively) in the discriminatory power of WV-2 during the transition from wet to dry season (April). Using Random Forest classifier, the classification scenarios investigated showed that i) the 8-bands of the WV-2 sensor achieved higher classification accuracy for the April date (transition from wet to dry season, senescence) compared to the March date (peak productivity season) ii) the WV-2 spectral configuration systematically outperformed the IKONOS sensor spectral configuration and iii) the multi-temporal approach (March and April combined) improved the discrimination of tress species and produced the highest overall accuracy results at 80.4%. Consistent with the interspecies-SAM procedure, the yellow (605 nm) band also showed a statistically significant contribution in the improved classification accuracy from WV-2. These results highlight the mapping opportunities presented by WV-2 data for monitoring the distribution status of e.g. species often harvested by local communities (e.g. Sclerocharya birrea), encroaching species, or species-specific tree losses induced by elephants.

The Accuracy and Reliability of Crowdsource Annotations of Digital Retinal Images.

PubMed

Mitry, Danny; Zutis, Kris; Dhillon, Baljean; Peto, Tunde; Hayat, Shabina; Khaw, Kay-Tee; Morgan, James E; Moncur, Wendy; Trucco, Emanuele; Foster, Paul J

2016-09-01

Crowdsourcing is based on outsourcing computationally intensive tasks to numerous individuals in the online community who have no formal training. Our aim was to develop a novel online tool designed to facilitate large-scale annotation of digital retinal images, and to assess the accuracy of crowdsource grading using this tool, comparing it to expert classification. We used 100 retinal fundus photograph images with predetermined disease criteria selected by two experts from a large cohort study. The Amazon Mechanical Turk Web platform was used to drive traffic to our site so anonymous workers could perform a classification and annotation task of the fundus photographs in our dataset after a short training exercise. Three groups were assessed: masters only, nonmasters only and nonmasters with compulsory training. We calculated the sensitivity, specificity, and area under the curve (AUC) of receiver operating characteristic (ROC) plots for all classifications compared to expert grading, and used the Dice coefficient and consensus threshold to assess annotation accuracy. In total, we received 5389 annotations for 84 images (excluding 16 training images) in 2 weeks. A specificity and sensitivity of 71% (95% confidence interval [CI], 69%-74%) and 87% (95% CI, 86%-88%) was achieved for all classifications. The AUC in this study for all classifications combined was 0.93 (95% CI, 0.91-0.96). For image annotation, a maximal Dice coefficient (∼0.6) was achieved with a consensus threshold of 0.25. This study supports the hypothesis that annotation of abnormalities in retinal images by ophthalmologically naive individuals is comparable to expert annotation. The highest AUC and agreement with expert annotation was achieved in the nonmasters with compulsory training group. The use of crowdsourcing as a technique for retinal image analysis may be comparable to expert graders and has the potential to deliver timely, accurate, and cost-effective image analysis.

Multivariate pattern analysis of obsessive-compulsive disorder using structural neuroanatomy.

PubMed

Hu, Xinyu; Liu, Qi; Li, Bin; Tang, Wanjie; Sun, Huaiqiang; Li, Fei; Yang, Yanchun; Gong, Qiyong; Huang, Xiaoqi

2016-02-01

Magnetic resonance imaging (MRI) studies have revealed brain structural abnormalities in obsessive-compulsive disorder (OCD) patients, involving both gray matter (GM) and white matter (WM). However, the results of previous publications were based on average differences between groups, which limited their usages in clinical practice. Therefore, the aim of this study was to examine whether the application of multivariate pattern analysis (MVPA) to high-dimensional structural images would allow accurate discrimination between OCD patients and healthy control subjects (HCS). High-resolution T1-weighted images were acquired from 33 OCD patients and 33 demographically matched HCS in a 3.0 T scanner. Differences in GM and WM volume between OCD and HCS were examined using two types of well-established MVPA techniques: support vector machine (SVM) and Gaussian process classifier (GPC). We also drew a receiver operating characteristic (ROC) curve to evaluate the performance of each classifier. The classification accuracies for both classifiers using GM and WM anatomy were all above 75%. The highest classification accuracy (81.82%, P<0.001) was achieved with the SVM classifier using WM information. Regional brain anomalies with high discriminative power were based on three distributed networks including the fronto-striatal circuit, the temporo-parieto-occipital junction and the cerebellum. Our study illustrated that both GM and WM anatomical features may be useful in differentiating OCD patients from HCS. WM volume using the SVM approach showed the highest accuracy in our population for revealing group differences, which suggested its potential diagnostic role in detecting highly enriched OCD patients at the level of the individual. Copyright © 2015 Elsevier B.V. and ECNP. All rights reserved.

Articular cartilage degeneration classification by means of high-frequency ultrasound.

PubMed

Männicke, N; Schöne, M; Oelze, M; Raum, K

2014-10-01

To date only single ultrasound parameters were regarded in statistical analyses to characterize osteoarthritic changes in articular cartilage and the potential benefit of using parameter combinations for characterization remains unclear. Therefore, the aim of this work was to utilize feature selection and classification of a Mankin subset score (i.e., cartilage surface and cell sub-scores) using ultrasound-based parameter pairs and investigate both classification accuracy and the sensitivity towards different degeneration stages. 40 punch biopsies of human cartilage were previously scanned ex vivo with a 40-MHz transducer. Ultrasound-based surface parameters, as well as backscatter and envelope statistics parameters were available. Logistic regression was performed with each unique US parameter pair as predictor and different degeneration stages as response variables. The best ultrasound-based parameter pair for each Mankin subset score value was assessed by highest classification accuracy and utilized in receiver operating characteristics (ROC) analysis. The classifications discriminating between early degenerations yielded area under the ROC curve (AUC) values of 0.94-0.99 (mean ± SD: 0.97 ± 0.03). In contrast, classifications among higher Mankin subset scores resulted in lower AUC values: 0.75-0.91 (mean ± SD: 0.84 ± 0.08). Variable sensitivities of the different ultrasound features were observed with respect to different degeneration stages. Our results strongly suggest that combinations of high-frequency ultrasound-based parameters exhibit potential to characterize different, particularly very early, degeneration stages of hyaline cartilage. Variable sensitivities towards different degeneration stages suggest that a concurrent estimation of multiple ultrasound-based parameters is diagnostically valuable. In-vivo application of the present findings is conceivable in both minimally invasive arthroscopic ultrasound and high-frequency transcutaneous ultrasound. Copyright © 2014 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved.

An ensemble learning system for a 4-way classification of Alzheimer's disease and mild cognitive impairment.

PubMed

Yao, Dongren; Calhoun, Vince D; Fu, Zening; Du, Yuhui; Sui, Jing

2018-05-15

Discriminating Alzheimer's disease (AD) from its prodromal form, mild cognitive impairment (MCI), is a significant clinical problem that may facilitate early diagnosis and intervention, in which a more challenging issue is to classify MCI subtypes, i.e., those who eventually convert to AD (cMCI) versus those who do not (MCI). To solve this difficult 4-way classification problem (AD, MCI, cMCI and healthy controls), a competition was hosted by Kaggle to invite the scientific community to apply their machine learning approaches on pre-processed sets of T1-weighted magnetic resonance images (MRI) data and the demographic information from the international Alzheimer's disease neuroimaging initiative (ADNI) database. This paper summarizes our competition results. We first proposed a hierarchical process by turning the 4-way classification into five binary classification problems. A new feature selection technology based on relative importance was also proposed, aiming to identify a more informative and concise subset from 426 sMRI morphometric and 3 demographic features, to ensure each binary classifier to achieve its highest accuracy. As a result, about 2% of the original features were selected to build a new feature space, which can achieve the final four-way classification with a 54.38% accuracy on testing data through hierarchical grouping, higher than several alternative methods in comparison. More importantly, the selected discriminative features such as hippocampal volume, parahippocampal surface area, and medial orbitofrontal thickness, etc. as well as the MMSE score, are reasonable and consistent with those reported in AD/MCI deficits. In summary, the proposed method provides a new framework for multi-way classification using hierarchical grouping and precise feature selection. Copyright © 2018 Elsevier B.V. All rights reserved.

A multi-label learning based kernel automatic recommendation method for support vector machine.

PubMed

Zhang, Xueying; Song, Qinbao

2015-01-01

Choosing an appropriate kernel is very important and critical when classifying a new problem with Support Vector Machine. So far, more attention has been paid on constructing new kernels and choosing suitable parameter values for a specific kernel function, but less on kernel selection. Furthermore, most of current kernel selection methods focus on seeking a best kernel with the highest classification accuracy via cross-validation, they are time consuming and ignore the differences among the number of support vectors and the CPU time of SVM with different kernels. Considering the tradeoff between classification success ratio and CPU time, there may be multiple kernel functions performing equally well on the same classification problem. Aiming to automatically select those appropriate kernel functions for a given data set, we propose a multi-label learning based kernel recommendation method built on the data characteristics. For each data set, the meta-knowledge data base is first created by extracting the feature vector of data characteristics and identifying the corresponding applicable kernel set. Then the kernel recommendation model is constructed on the generated meta-knowledge data base with the multi-label classification method. Finally, the appropriate kernel functions are recommended to a new data set by the recommendation model according to the characteristics of the new data set. Extensive experiments over 132 UCI benchmark data sets, with five different types of data set characteristics, eleven typical kernels (Linear, Polynomial, Radial Basis Function, Sigmoidal function, Laplace, Multiquadric, Rational Quadratic, Spherical, Spline, Wave and Circular), and five multi-label classification methods demonstrate that, compared with the existing kernel selection methods and the most widely used RBF kernel function, SVM with the kernel function recommended by our proposed method achieved the highest classification performance.

A Multi-Label Learning Based Kernel Automatic Recommendation Method for Support Vector Machine

PubMed Central

Zhang, Xueying; Song, Qinbao

2015-01-01

Choosing an appropriate kernel is very important and critical when classifying a new problem with Support Vector Machine. So far, more attention has been paid on constructing new kernels and choosing suitable parameter values for a specific kernel function, but less on kernel selection. Furthermore, most of current kernel selection methods focus on seeking a best kernel with the highest classification accuracy via cross-validation, they are time consuming and ignore the differences among the number of support vectors and the CPU time of SVM with different kernels. Considering the tradeoff between classification success ratio and CPU time, there may be multiple kernel functions performing equally well on the same classification problem. Aiming to automatically select those appropriate kernel functions for a given data set, we propose a multi-label learning based kernel recommendation method built on the data characteristics. For each data set, the meta-knowledge data base is first created by extracting the feature vector of data characteristics and identifying the corresponding applicable kernel set. Then the kernel recommendation model is constructed on the generated meta-knowledge data base with the multi-label classification method. Finally, the appropriate kernel functions are recommended to a new data set by the recommendation model according to the characteristics of the new data set. Extensive experiments over 132 UCI benchmark data sets, with five different types of data set characteristics, eleven typical kernels (Linear, Polynomial, Radial Basis Function, Sigmoidal function, Laplace, Multiquadric, Rational Quadratic, Spherical, Spline, Wave and Circular), and five multi-label classification methods demonstrate that, compared with the existing kernel selection methods and the most widely used RBF kernel function, SVM with the kernel function recommended by our proposed method achieved the highest classification performance. PMID:25893896

Genetic Bee Colony (GBC) algorithm: A new gene selection method for microarray cancer classification.

PubMed

Alshamlan, Hala M; Badr, Ghada H; Alohali, Yousef A

2015-06-01

Naturally inspired evolutionary algorithms prove effectiveness when used for solving feature selection and classification problems. Artificial Bee Colony (ABC) is a relatively new swarm intelligence method. In this paper, we propose a new hybrid gene selection method, namely Genetic Bee Colony (GBC) algorithm. The proposed algorithm combines the used of a Genetic Algorithm (GA) along with Artificial Bee Colony (ABC) algorithm. The goal is to integrate the advantages of both algorithms. The proposed algorithm is applied to a microarray gene expression profile in order to select the most predictive and informative genes for cancer classification. In order to test the accuracy performance of the proposed algorithm, extensive experiments were conducted. Three binary microarray datasets are use, which include: colon, leukemia, and lung. In addition, another three multi-class microarray datasets are used, which are: SRBCT, lymphoma, and leukemia. Results of the GBC algorithm are compared with our recently proposed technique: mRMR when combined with the Artificial Bee Colony algorithm (mRMR-ABC). We also compared the combination of mRMR with GA (mRMR-GA) and Particle Swarm Optimization (mRMR-PSO) algorithms. In addition, we compared the GBC algorithm with other related algorithms that have been recently published in the literature, using all benchmark datasets. The GBC algorithm shows superior performance as it achieved the highest classification accuracy along with the lowest average number of selected genes. This proves that the GBC algorithm is a promising approach for solving the gene selection problem in both binary and multi-class cancer classification. Copyright © 2015 Elsevier Ltd. All rights reserved.

[Accuracy improvement of spectral classification of crop using microwave backscatter data].

PubMed

Jia, Kun; Li, Qiang-Zi; Tian, Yi-Chen; Wu, Bing-Fang; Zhang, Fei-Fei; Meng, Ji-Hua

2011-02-01

In the present study, VV polarization microwave backscatter data used for improving accuracies of spectral classification of crop is investigated. Classification accuracy using different classifiers based on the fusion data of HJ satellite multi-spectral and Envisat ASAR VV backscatter data are compared. The results indicate that fusion data can take full advantage of spectral information of HJ multi-spectral data and the structure sensitivity feature of ASAR VV polarization data. The fusion data enlarges the spectral difference among different classifications and improves crop classification accuracy. The classification accuracy using fusion data can be increased by 5 percent compared to the single HJ data. Furthermore, ASAR VV polarization data is sensitive to non-agrarian area of planted field, and VV polarization data joined classification can effectively distinguish the field border. VV polarization data associating with multi-spectral data used in crop classification enlarges the application of satellite data and has the potential of spread in the domain of agriculture.

Semi-automatic classification of glaciovolcanic landforms: An object-based mapping approach based on geomorphometry

NASA Astrophysics Data System (ADS)

Pedersen, G. B. M.

2016-02-01

A new object-oriented approach is developed to classify glaciovolcanic landforms (Procedure A) and their landform elements boundaries (Procedure B). It utilizes the principle that glaciovolcanic edifices are geomorphometrically distinct from lava shields and plains (Pedersen and Grosse, 2014), and the approach is tested on data from Reykjanes Peninsula, Iceland. The outlined procedures utilize slope and profile curvature attribute maps (20 m/pixel) and the classified results are evaluated quantitatively through error matrix maps (Procedure A) and visual inspection (Procedure B). In procedure A, the highest obtained accuracy is 94.1%, but even simple mapping procedures provide good results (> 90% accuracy). Successful classification of glaciovolcanic landform element boundaries (Procedure B) is also achieved and this technique has the potential to delineate the transition from intraglacial to subaerial volcanic activity in orthographic view. This object-oriented approach based on geomorphometry overcomes issues with vegetation cover, which has been typically problematic for classification schemes utilizing spectral data. Furthermore, it handles complex edifice outlines well and is easily incorporated into a GIS environment, where results can be edited or fused with other mapping results. The approach outlined here is designed to map glaciovolcanic edifices within the Icelandic neovolcanic zone but may also be applied to similar subaerial or submarine volcanic settings, where steep volcanic edifices are surrounded by flat plains.

Differential Memory Test Sensitivity for Diagnosing Amnestic Mild Cognitive Impairment and Predicting Conversion to Alzheimer's Disease

PubMed Central

Rabin, Laura A.; Paré, Nadia; Saykin, Andrew J.; Brown, Michael J.; Wishart, Heather A.; Flashman, Laura A.; Santulli, Robert B.

2011-01-01

Episodic memory is the first and most severely affected cognitive domain in Alzheimer's disease (AD), and it is also the key early marker in prodromal stages including amnestic mild cognitive impairment (MCI). The relative ability of memory tests to discriminate between MCI and normal aging has not been well characterized. We compared the classification value of widely used verbal memory tests in distinguishing healthy older adults (n = 51) from those with MCI (n = 38). Univariate logistic regression indicated that the total learning score from the California Verbal Learning Test-II (CVLT-II) ranked highest in terms of distinguishing MCI from normal aging (sensitivity = 90.2; specificity = 84.2). Inclusion of the delayed recall condition of a story memory task (i.e., WMS-III Logical Memory, Story A) enhanced the overall accuracy of classification (sensitivity = 92.2; specificity = 94.7). Combining Logical Memory recognition and CVLT-II long delay best predicted progression from MCI to AD over a 4-year period (accurate classification = 87.5%). Learning across multiple trials may provide the most sensitive index for initial diagnosis of MCI, but inclusion of additional variables may enhance overall accuracy and may represent the optimal strategy for identifying individuals most likely to progress to dementia. PMID:19353345

Combining features from ERP components in single-trial EEG for discriminating four-category visual objects.

PubMed

Wang, Changming; Xiong, Shi; Hu, Xiaoping; Yao, Li; Zhang, Jiacai

2012-10-01

Categorization of images containing visual objects can be successfully recognized using single-trial electroencephalograph (EEG) measured when subjects view images. Previous studies have shown that task-related information contained in event-related potential (ERP) components could discriminate two or three categories of object images. In this study, we investigated whether four categories of objects (human faces, buildings, cats and cars) could be mutually discriminated using single-trial EEG data. Here, the EEG waveforms acquired while subjects were viewing four categories of object images were segmented into several ERP components (P1, N1, P2a and P2b), and then Fisher linear discriminant analysis (Fisher-LDA) was used to classify EEG features extracted from ERP components. Firstly, we compared the classification results using features from single ERP components, and identified that the N1 component achieved the highest classification accuracies. Secondly, we discriminated four categories of objects using combining features from multiple ERP components, and showed that combination of ERP components improved four-category classification accuracies by utilizing the complementarity of discriminative information in ERP components. These findings confirmed that four categories of object images could be discriminated with single-trial EEG and could direct us to select effective EEG features for classifying visual objects.

A new automated quantification algorithm for the detection and evaluation of focal liver lesions with contrast-enhanced ultrasound.

PubMed

Gatos, Ilias; Tsantis, Stavros; Spiliopoulos, Stavros; Skouroliakou, Aikaterini; Theotokas, Ioannis; Zoumpoulis, Pavlos; Hazle, John D; Kagadis, George C

2015-07-01

Detect and classify focal liver lesions (FLLs) from contrast-enhanced ultrasound (CEUS) imaging by means of an automated quantification algorithm. The proposed algorithm employs a sophisticated segmentation method to detect and contour focal lesions from 52 CEUS video sequences (30 benign and 22 malignant). Lesion detection involves wavelet transform zero crossings utilization as an initialization step to the Markov random field model toward the lesion contour extraction. After FLL detection across frames, time intensity curve (TIC) is computed which provides the contrast agents' behavior at all vascular phases with respect to adjacent parenchyma for each patient. From each TIC, eight features were automatically calculated and employed into the support vector machines (SVMs) classification algorithm in the design of the image analysis model. With regard to FLLs detection accuracy, all lesions detected had an average overlap value of 0.89 ± 0.16 with manual segmentations for all CEUS frame-subsets included in the study. Highest classification accuracy from the SVM model was 90.3%, misdiagnosing three benign and two malignant FLLs with sensitivity and specificity values of 93.1% and 86.9%, respectively. The proposed quantification system that employs FLLs detection and classification algorithms may be of value to physicians as a second opinion tool for avoiding unnecessary invasive procedures.

High blood Pressure in children and its correlation with three definitions of obesity in childhood

PubMed Central

de Moraes, Leonardo Iezzi; Nicola, Thaís Coutinho; de Jesus, Julyanna Silva Araújo; Alves, Eduardo Roberty Badiani; Giovaninni, Nayara Paula Bernurdes; Marcato, Daniele Gasparini; Sampaio, Jéssica Dutra; Fuly, Jeanne Teixeira Bessa; Costalonga, Everlayny Fiorot

2014-01-01

Background Several authors have correlated the increase of cardiovascular risk with the nutritional status, however there are different criteria for the classification of overweight and obesity in children. Objectives To evaluate the performance of three nutritional classification criteria in children, as definers of the presence of obesity and predictors of high blood pressure in schoolchildren. Methods Eight hundred and seventeen children ranging 6 to 13 years old, enrolled in public schools in the municipality of Vila Velha (ES) were submitted to anthropometric evaluation and blood pressure measurement. The classification of the nutritional status was established by two international criteria (CDC/NCHS 2000 and IOTF 2000) and one Brazilian criterion (Conde e Monteiro 2006). Results The prevalence of overweight was higher when the criterion of Conde e Monteiro (27%) was used, and inferior by the IOTF (15%) criteria. High blood pressure was observed in 7.3% of children. It was identified a strong association between the presence of overweight and the occurrence of high blood pressure, regardless of the test used (p < 0.001). The test showing the highest sensitivity in predicting elevated BP was the Conde e Monteiro (44%), while the highest specificity (94%) and greater overall accuracy (63%), was the CDC criterion. Conclusions The prevalence of overweight in Brazilian children is higher when using the classification criterion of Conde e Monteiro, and lower when the criterion used is IOTF. The Brazilian classification criterion proved to be the most sensitive predictor of high BP risk in this sample. PMID:24676372

Multi-Feature Classification of Multi-Sensor Satellite Imagery Based on Dual-Polarimetric Sentinel-1A, Landsat-8 OLI, and Hyperion Images for Urban Land-Cover Classification.

PubMed

Zhou, Tao; Li, Zhaofu; Pan, Jianjun

2018-01-27

This paper focuses on evaluating the ability and contribution of using backscatter intensity, texture, coherence, and color features extracted from Sentinel-1A data for urban land cover classification and comparing different multi-sensor land cover mapping methods to improve classification accuracy. Both Landsat-8 OLI and Hyperion images were also acquired, in combination with Sentinel-1A data, to explore the potential of different multi-sensor urban land cover mapping methods to improve classification accuracy. The classification was performed using a random forest (RF) method. The results showed that the optimal window size of the combination of all texture features was 9 × 9, and the optimal window size was different for each individual texture feature. For the four different feature types, the texture features contributed the most to the classification, followed by the coherence and backscatter intensity features; and the color features had the least impact on the urban land cover classification. Satisfactory classification results can be obtained using only the combination of texture and coherence features, with an overall accuracy up to 91.55% and a kappa coefficient up to 0.8935, respectively. Among all combinations of Sentinel-1A-derived features, the combination of the four features had the best classification result. Multi-sensor urban land cover mapping obtained higher classification accuracy. The combination of Sentinel-1A and Hyperion data achieved higher classification accuracy compared to the combination of Sentinel-1A and Landsat-8 OLI images, with an overall accuracy of up to 99.12% and a kappa coefficient up to 0.9889. When Sentinel-1A data was added to Hyperion images, the overall accuracy and kappa coefficient were increased by 4.01% and 0.0519, respectively.

Rescuing the Clinical Breast Examination: Advances in Classifying Technique and Assessing Physician Competency.

PubMed

Laufer, Shlomi; D'Angelo, Anne-Lise D; Kwan, Calvin; Ray, Rebbeca D; Yudkowsky, Rachel; Boulet, John R; McGaghie, William C; Pugh, Carla M

2017-12-01

Develop new performance evaluation standards for the clinical breast examination (CBE). There are several, technical aspects of a proper CBE. Our recent work discovered a significant, linear relationship between palpation force and CBE accuracy. This article investigates the relationship between other technical aspects of the CBE and accuracy. This performance assessment study involved data collection from physicians (n = 553) attending 3 different clinical meetings between 2013 and 2014: American Society of Breast Surgeons, American Academy of Family Physicians, and American College of Obstetricians and Gynecologists. Four, previously validated, sensor-enabled breast models were used for clinical skills assessment. Models A and B had solitary, superficial, 2 cm and 1 cm soft masses, respectively. Models C and D had solitary, deep, 2 cm hard and moderately firm masses, respectively. Finger movements (search technique) from 1137 CBE video recordings were independently classified by 2 observers. Final classifications were compared with CBE accuracy. Accuracy rates were model A = 99.6%, model B = 89.7%, model C = 75%, and model D = 60%. Final classification categories for search technique included rubbing movement, vertical movement, piano fingers, and other. Interrater reliability was (k = 0.79). Rubbing movement was 4 times more likely to yield an accurate assessment (odds ratio 3.81, P < 0.001) compared with vertical movement and piano fingers. Piano fingers had the highest failure rate (36.5%). Regression analysis of search pattern, search technique, palpation force, examination time, and 6 demographic variables, revealed that search technique independently and significantly affected CBE accuracy (P < 0.001). Our results support measurement and classification of CBE techniques and provide the foundation for a new paradigm in teaching and assessing hands-on clinical skills. The newly described piano fingers palpation technique was noted to have unusually high failure rates. Medical educators should be aware of the potential differences in effectiveness for various CBE techniques.

Can segmentation evaluation metric be used as an indicator of land cover classification accuracy?

NASA Astrophysics Data System (ADS)

Švab Lenarčič, Andreja; Đurić, Nataša; Čotar, Klemen; Ritlop, Klemen; Oštir, Krištof

2016-10-01

It is a broadly established belief that the segmentation result significantly affects subsequent image classification accuracy. However, the actual correlation between the two has never been evaluated. Such an evaluation would be of considerable importance for any attempts to automate the object-based classification process, as it would reduce the amount of user intervention required to fine-tune the segmentation parameters. We conducted an assessment of segmentation and classification by analyzing 100 different segmentation parameter combinations, 3 classifiers, 5 land cover classes, 20 segmentation evaluation metrics, and 7 classification accuracy measures. The reliability definition of segmentation evaluation metrics as indicators of land cover classification accuracy was based on the linear correlation between the two. All unsupervised metrics that are not based on number of segments have a very strong correlation with all classification measures and are therefore reliable as indicators of land cover classification accuracy. On the other hand, correlation at supervised metrics is dependent on so many factors that it cannot be trusted as a reliable classification quality indicator. Algorithms for land cover classification studied in this paper are widely used; therefore, presented results are applicable to a wider area.

Combined texture feature analysis of segmentation and classification of benign and malignant tumour CT slices.

PubMed

Padma, A; Sukanesh, R

2013-01-01

A computer software system is designed for the segmentation and classification of benign from malignant tumour slices in brain computed tomography (CT) images. This paper presents a method to find and select both the dominant run length and co-occurrence texture features of region of interest (ROI) of the tumour region of each slice to be segmented by Fuzzy c means clustering (FCM) and evaluate the performance of support vector machine (SVM)-based classifiers in classifying benign and malignant tumour slices. Two hundred and six tumour confirmed CT slices are considered in this study. A total of 17 texture features are extracted by a feature extraction procedure, and six features are selected using Principal Component Analysis (PCA). This study constructed the SVM-based classifier with the selected features and by comparing the segmentation results with the experienced radiologist labelled ground truth (target). Quantitative analysis between ground truth and segmented tumour is presented in terms of segmentation accuracy, segmentation error and overlap similarity measures such as the Jaccard index. The classification performance of the SVM-based classifier with the same selected features is also evaluated using a 10-fold cross-validation method. The proposed system provides some newly found texture features have an important contribution in classifying benign and malignant tumour slices efficiently and accurately with less computational time. The experimental results showed that the proposed system is able to achieve the highest segmentation and classification accuracy effectiveness as measured by jaccard index and sensitivity and specificity.

Quantifying and Characterizing Tonic Thermal Pain Across Subjects From EEG Data Using Random Forest Models.

PubMed

Vijayakumar, Vishal; Case, Michelle; Shirinpour, Sina; He, Bin

2017-12-01

Effective pain assessment and management strategies are needed to better manage pain. In addition to self-report, an objective pain assessment system can provide a more complete picture of the neurophysiological basis for pain. In this study, a robust and accurate machine learning approach is developed to quantify tonic thermal pain across healthy subjects into a maximum of ten distinct classes. A random forest model was trained to predict pain scores using time-frequency wavelet representations of independent components obtained from electroencephalography (EEG) data, and the relative importance of each frequency band to pain quantification is assessed. The mean classification accuracy for predicting pain on an independent test subject for a range of 1-10 is 89.45%, highest among existing state of the art quantification algorithms for EEG. The gamma band is the most important to both intersubject and intrasubject classification accuracy. The robustness and generalizability of the classifier are demonstrated. Our results demonstrate the potential of this tool to be used clinically to help us to improve chronic pain treatment and establish spectral biomarkers for future pain-related studies using EEG.

Classifying Radio Galaxies with the Convolutional Neural Network

NASA Astrophysics Data System (ADS)

Aniyan, A. K.; Thorat, K.

2017-06-01

We present the application of a deep machine learning technique to classify radio images of extended sources on a morphological basis using convolutional neural networks (CNN). In this study, we have taken the case of the Fanaroff-Riley (FR) class of radio galaxies as well as radio galaxies with bent-tailed morphology. We have used archival data from the Very Large Array (VLA)—Faint Images of the Radio Sky at Twenty Centimeters survey and existing visually classified samples available in the literature to train a neural network for morphological classification of these categories of radio sources. Our training sample size for each of these categories is ˜200 sources, which has been augmented by rotated versions of the same. Our study shows that CNNs can classify images of the FRI and FRII and bent-tailed radio galaxies with high accuracy (maximum precision at 95%) using well-defined samples and a “fusion classifier,” which combines the results of binary classifications, while allowing for a mechanism to find sources with unusual morphologies. The individual precision is highest for bent-tailed radio galaxies at 95% and is 91% and 75% for the FRI and FRII classes, respectively, whereas the recall is highest for FRI and FRIIs at 91% each, while the bent-tailed class has a recall of 79%. These results show that our results are comparable to that of manual classification, while being much faster. Finally, we discuss the computational and data-related challenges associated with the morphological classification of radio galaxies with CNNs.

Decision support system for diabetic retinopathy using discrete wavelet transform.

PubMed

Noronha, K; Acharya, U R; Nayak, K P; Kamath, S; Bhandary, S V

2013-03-01

Prolonged duration of the diabetes may affect the tiny blood vessels of the retina causing diabetic retinopathy. Routine eye screening of patients with diabetes helps to detect diabetic retinopathy at the early stage. It is very laborious and time-consuming for the doctors to go through many fundus images continuously. Therefore, decision support system for diabetic retinopathy detection can reduce the burden of the ophthalmologists. In this work, we have used discrete wavelet transform and support vector machine classifier for automated detection of normal and diabetic retinopathy classes. The wavelet-based decomposition was performed up to the second level, and eight energy features were extracted. Two energy features from the approximation coefficients of two levels and six energy values from the details in three orientations (horizontal, vertical and diagonal) were evaluated. These features were fed to the support vector machine classifier with various kernel functions (linear, radial basis function, polynomial of orders 2 and 3) to evaluate the highest classification accuracy. We obtained the highest average classification accuracy, sensitivity and specificity of more than 99% with support vector machine classifier (polynomial kernel of order 3) using three discrete wavelet transform features. We have also proposed an integrated index called Diabetic Retinopathy Risk Index using clinically significant wavelet energy features to identify normal and diabetic retinopathy classes using just one number. We believe that this (Diabetic Retinopathy Risk Index) can be used as an adjunct tool by the doctors during the eye screening to cross-check their diagnosis.

A new computer aided diagnosis system for evaluation of chronic liver disease with ultrasound shear wave elastography imaging.

PubMed

Gatos, Ilias; Tsantis, Stavros; Spiliopoulos, Stavros; Karnabatidis, Dimitris; Theotokas, Ioannis; Zoumpoulis, Pavlos; Loupas, Thanasis; Hazle, John D; Kagadis, George C

2016-03-01

Classify chronic liver disease (CLD) from ultrasound shear-wave elastography (SWE) imaging by means of a computer aided diagnosis (CAD) system. The proposed algorithm employs an inverse mapping technique (red-green-blue to stiffness) to quantify 85 SWE images (54 healthy and 31 with CLD). Texture analysis is then applied involving the automatic calculation of 330 first and second order textural features from every transformed stiffness value map to determine functional features that characterize liver elasticity and describe liver condition for all available stages. Consequently, a stepwise regression analysis feature selection procedure is utilized toward a reduced feature subset that is fed into the support vector machines (SVMs) classification algorithm in the design of the CAD system. With regard to the mapping procedure accuracy, the stiffness map values had an average difference of 0.01 ± 0.001 kPa compared to the quantification results derived from the color-box provided by the built-in software of the ultrasound system. Highest classification accuracy from the SVM model was 87.0% with sensitivity and specificity values of 83.3% and 89.1%, respectively. Receiver operating characteristic curves analysis gave an area under the curve value of 0.85 with [0.77-0.89] confidence interval. The proposed CAD system employing color to stiffness mapping and classification algorithms offered superior results, comparing the already published clinical studies. It could prove to be of value to physicians improving the diagnostic accuracy of CLD and can be employed as a second opinion tool for avoiding unnecessary invasive procedures.

Posture Detection Based on Smart Cushion for Wheelchair Users

PubMed Central

Ma, Congcong; Li, Wenfeng; Gravina, Raffaele; Fortino, Giancarlo

2017-01-01

The postures of wheelchair users can reveal their sitting habit, mood, and even predict health risks such as pressure ulcers or lower back pain. Mining the hidden information of the postures can reveal their wellness and general health conditions. In this paper, a cushion-based posture recognition system is used to process pressure sensor signals for the detection of user’s posture in the wheelchair. The proposed posture detection method is composed of three main steps: data level classification for posture detection, backward selection of sensor configuration, and recognition results compared with previous literature. Five supervised classification techniques—Decision Tree (J48), Support Vector Machines (SVM), Multilayer Perceptron (MLP), Naive Bayes, and k-Nearest Neighbor (k-NN)—are compared in terms of classification accuracy, precision, recall, and F-measure. Results indicate that the J48 classifier provides the highest accuracy compared to other techniques. The backward selection method was used to determine the best sensor deployment configuration of the wheelchair. Several kinds of pressure sensor deployments are compared and our new method of deployment is shown to better detect postures of the wheelchair users. Performance analysis also took into account the Body Mass Index (BMI), useful for evaluating the robustness of the method across individual physical differences. Results show that our proposed sensor deployment is effective, achieving 99.47% posture recognition accuracy. Our proposed method is very competitive for posture recognition and robust in comparison with other former research. Accurate posture detection represents a fundamental basic block to develop several applications, including fatigue estimation and activity level assessment. PMID:28353684

Land-cover classification in a moist tropical region of Brazil with Landsat TM imagery.

PubMed

Li, Guiying; Lu, Dengsheng; Moran, Emilio; Hetrick, Scott

2011-01-01

This research aims to improve land-cover classification accuracy in a moist tropical region in Brazil by examining the use of different remote sensing-derived variables and classification algorithms. Different scenarios based on Landsat Thematic Mapper (TM) spectral data and derived vegetation indices and textural images, and different classification algorithms - maximum likelihood classification (MLC), artificial neural network (ANN), classification tree analysis (CTA), and object-based classification (OBC), were explored. The results indicated that a combination of vegetation indices as extra bands into Landsat TM multispectral bands did not improve the overall classification performance, but the combination of textural images was valuable for improving vegetation classification accuracy. In particular, the combination of both vegetation indices and textural images into TM multispectral bands improved overall classification accuracy by 5.6% and kappa coefficient by 6.25%. Comparison of the different classification algorithms indicated that CTA and ANN have poor classification performance in this research, but OBC improved primary forest and pasture classification accuracies. This research indicates that use of textural images or use of OBC are especially valuable for improving the vegetation classes such as upland and liana forest classes having complex stand structures and having relatively large patch sizes.

Land-cover classification in a moist tropical region of Brazil with Landsat TM imagery

PubMed Central

LI, GUIYING; LU, DENGSHENG; MORAN, EMILIO; HETRICK, SCOTT

2011-01-01

This research aims to improve land-cover classification accuracy in a moist tropical region in Brazil by examining the use of different remote sensing-derived variables and classification algorithms. Different scenarios based on Landsat Thematic Mapper (TM) spectral data and derived vegetation indices and textural images, and different classification algorithms – maximum likelihood classification (MLC), artificial neural network (ANN), classification tree analysis (CTA), and object-based classification (OBC), were explored. The results indicated that a combination of vegetation indices as extra bands into Landsat TM multispectral bands did not improve the overall classification performance, but the combination of textural images was valuable for improving vegetation classification accuracy. In particular, the combination of both vegetation indices and textural images into TM multispectral bands improved overall classification accuracy by 5.6% and kappa coefficient by 6.25%. Comparison of the different classification algorithms indicated that CTA and ANN have poor classification performance in this research, but OBC improved primary forest and pasture classification accuracies. This research indicates that use of textural images or use of OBC are especially valuable for improving the vegetation classes such as upland and liana forest classes having complex stand structures and having relatively large patch sizes. PMID:22368311

Use of Binary Partition Tree and energy minimization for object-based classification of urban land cover

NASA Astrophysics Data System (ADS)

Li, Mengmeng; Bijker, Wietske; Stein, Alfred

2015-04-01

Two main challenges are faced when classifying urban land cover from very high resolution satellite images: obtaining an optimal image segmentation and distinguishing buildings from other man-made objects. For optimal segmentation, this work proposes a hierarchical representation of an image by means of a Binary Partition Tree (BPT) and an unsupervised evaluation of image segmentations by energy minimization. For building extraction, we apply fuzzy sets to create a fuzzy landscape of shadows which in turn involves a two-step procedure. The first step is a preliminarily image classification at a fine segmentation level to generate vegetation and shadow information. The second step models the directional relationship between building and shadow objects to extract building information at the optimal segmentation level. We conducted the experiments on two datasets of Pléiades images from Wuhan City, China. To demonstrate its performance, the proposed classification is compared at the optimal segmentation level with Maximum Likelihood Classification and Support Vector Machine classification. The results show that the proposed classification produced the highest overall accuracies and kappa coefficients, and the smallest over-classification and under-classification geometric errors. We conclude first that integrating BPT with energy minimization offers an effective means for image segmentation. Second, we conclude that the directional relationship between building and shadow objects represented by a fuzzy landscape is important for building extraction.

Classification of right-hand grasp movement based on EMOTIV Epoc+

NASA Astrophysics Data System (ADS)

Tobing, T. A. M. L.; Prawito, Wijaya, S. K.

2017-07-01

Combinations of BCT elements for right-hand grasp movement have been obtained, providing the average value of their classification accuracy. The aim of this study is to find a suitable combination for best classification accuracy of right-hand grasp movement based on EEG headset, EMOTIV Epoc+. There are three movement classifications: grasping hand, relax, and opening hand. These classifications take advantage of Event-Related Desynchronization (ERD) phenomenon that makes it possible to differ relaxation, imagery, and movement state from each other. The combinations of elements are the usage of Independent Component Analysis (ICA), spectrum analysis by Fast Fourier Transform (FFT), maximum mu and beta power with their frequency as features, and also classifier Probabilistic Neural Network (PNN) and Radial Basis Function (RBF). The average values of classification accuracy are ± 83% for training and ± 57% for testing. To have a better understanding of the signal quality recorded by EMOTIV Epoc+, the result of classification accuracy of left or right-hand grasping movement EEG signal (provided by Physionet) also be given, i.e.± 85% for training and ± 70% for testing. The comparison of accuracy value from each combination, experiment condition, and external EEG data are provided for the purpose of value analysis of classification accuracy.

Classification of savanna tree species, in the Greater Kruger National Park region, by integrating hyperspectral and LiDAR data in a Random Forest data mining environment

NASA Astrophysics Data System (ADS)

Naidoo, L.; Cho, M. A.; Mathieu, R.; Asner, G.

2012-04-01

The accurate classification and mapping of individual trees at species level in the savanna ecosystem can provide numerous benefits for the managerial authorities. Such benefits include the mapping of economically useful tree species, which are a key source of food production and fuel wood for the local communities, and of problematic alien invasive and bush encroaching species, which can threaten the integrity of the environment and livelihoods of the local communities. Species level mapping is particularly challenging in African savannas which are complex, heterogeneous, and open environments with high intra-species spectral variability due to differences in geology, topography, rainfall, herbivory and human impacts within relatively short distances. Savanna vegetation are also highly irregular in canopy and crown shape, height and other structural dimensions with a combination of open grassland patches and dense woody thicket - a stark contrast to the more homogeneous forest vegetation. This study classified eight common savanna tree species in the Greater Kruger National Park region, South Africa, using a combination of hyperspectral and Light Detection and Ranging (LiDAR)-derived structural parameters, in the form of seven predictor datasets, in an automated Random Forest modelling approach. The most important predictors, which were found to play an important role in the different classification models and contributed to the success of the hybrid dataset model when combined, were species tree height; NDVI; the chlorophyll b wavelength (466 nm) and a selection of raw, continuum removed and Spectral Angle Mapper (SAM) bands. It was also concluded that the hybrid predictor dataset Random Forest model yielded the highest classification accuracy and prediction success for the eight savanna tree species with an overall classification accuracy of 87.68% and KHAT value of 0.843.

An integrative machine learning strategy for improved prediction of essential genes in Escherichia coli metabolism using flux-coupled features.

PubMed

Nandi, Sutanu; Subramanian, Abhishek; Sarkar, Ram Rup

2017-07-25

Prediction of essential genes helps to identify a minimal set of genes that are absolutely required for the appropriate functioning and survival of a cell. The available machine learning techniques for essential gene prediction have inherent problems, like imbalanced provision of training datasets, biased choice of the best model for a given balanced dataset, choice of a complex machine learning algorithm, and data-based automated selection of biologically relevant features for classification. Here, we propose a simple support vector machine-based learning strategy for the prediction of essential genes in Escherichia coli K-12 MG1655 metabolism that integrates a non-conventional combination of an appropriate sample balanced training set, a unique organism-specific genotype, phenotype attributes that characterize essential genes, and optimal parameters of the learning algorithm to generate the best machine learning model (the model with the highest accuracy among all the models trained for different sample training sets). For the first time, we also introduce flux-coupled metabolic subnetwork-based features for enhancing the classification performance. Our strategy proves to be superior as compared to previous SVM-based strategies in obtaining a biologically relevant classification of genes with high sensitivity and specificity. This methodology was also trained with datasets of other recent supervised classification techniques for essential gene classification and tested using reported test datasets. The testing accuracy was always high as compared to the known techniques, proving that our method outperforms known methods. Observations from our study indicate that essential genes are conserved among homologous bacterial species, demonstrate high codon usage bias, GC content and gene expression, and predominantly possess a tendency to form physiological flux modules in metabolism.

Hardware friendly probabilistic spiking neural network with long-term and short-term plasticity.

PubMed

Hsieh, Hung-Yi; Tang, Kea-Tiong

2013-12-01

This paper proposes a probabilistic spiking neural network (PSNN) with unimodal weight distribution, possessing long- and short-term plasticity. The proposed algorithm is derived by both the arithmetic gradient decent calculation and bioinspired algorithms. The algorithm is benchmarked by the Iris and Wisconsin breast cancer (WBC) data sets. The network features fast convergence speed and high accuracy. In the experiment, the PSNN took not more than 40 epochs for convergence. The average testing accuracy for Iris and WBC data is 96.7% and 97.2%, respectively. To test the usefulness of the PSNN to real world application, the PSNN was also tested with the odor data, which was collected by our self-developed electronic nose (e-nose). Compared with the algorithm (K-nearest neighbor) that has the highest classification accuracy in the e-nose for the same odor data, the classification accuracy of the PSNN is only 1.3% less but the memory requirement can be reduced at least 40%. All the experiments suggest that the PSNN is hardware friendly. First, it requires only nine-bits weight resolution for training and testing. Second, the PSNN can learn complex data sets with a little number of neurons that in turn reduce the cost of VLSI implementation. In addition, the algorithm is insensitive to synaptic noise and the parameter variation induced by the VLSI fabrication. Therefore, the algorithm can be implemented by either software or hardware, making it suitable for wider application.

Multi-Feature Classification of Multi-Sensor Satellite Imagery Based on Dual-Polarimetric Sentinel-1A, Landsat-8 OLI, and Hyperion Images for Urban Land-Cover Classification

PubMed Central

Pan, Jianjun

2018-01-01

This paper focuses on evaluating the ability and contribution of using backscatter intensity, texture, coherence, and color features extracted from Sentinel-1A data for urban land cover classification and comparing different multi-sensor land cover mapping methods to improve classification accuracy. Both Landsat-8 OLI and Hyperion images were also acquired, in combination with Sentinel-1A data, to explore the potential of different multi-sensor urban land cover mapping methods to improve classification accuracy. The classification was performed using a random forest (RF) method. The results showed that the optimal window size of the combination of all texture features was 9 × 9, and the optimal window size was different for each individual texture feature. For the four different feature types, the texture features contributed the most to the classification, followed by the coherence and backscatter intensity features; and the color features had the least impact on the urban land cover classification. Satisfactory classification results can be obtained using only the combination of texture and coherence features, with an overall accuracy up to 91.55% and a kappa coefficient up to 0.8935, respectively. Among all combinations of Sentinel-1A-derived features, the combination of the four features had the best classification result. Multi-sensor urban land cover mapping obtained higher classification accuracy. The combination of Sentinel-1A and Hyperion data achieved higher classification accuracy compared to the combination of Sentinel-1A and Landsat-8 OLI images, with an overall accuracy of up to 99.12% and a kappa coefficient up to 0.9889. When Sentinel-1A data was added to Hyperion images, the overall accuracy and kappa coefficient were increased by 4.01% and 0.0519, respectively. PMID:29382073

Estimating Classification Consistency and Accuracy for Cognitive Diagnostic Assessment

ERIC Educational Resources Information Center

Cui, Ying; Gierl, Mark J.; Chang, Hua-Hua

2012-01-01

This article introduces procedures for the computation and asymptotic statistical inference for classification consistency and accuracy indices specifically designed for cognitive diagnostic assessments. The new classification indices can be used as important indicators of the reliability and validity of classification results produced by…

The effect of finite field size on classification and atmospheric correction

NASA Technical Reports Server (NTRS)

Kaufman, Y. J.; Fraser, R. S.

1981-01-01

The atmospheric effect on the upward radiance of sunlight scattered from the Earth-atmosphere system is strongly influenced by the contrasts between fields and their sizes. For a given atmospheric turbidity, the atmospheric effect on classification of surface features is much stronger for nonuniform surfaces than for uniform surfaces. Therefore, the classification accuracy of agricultural fields and urban areas is dependent not only on the optical characteristics of the atmosphere, but also on the size of the surface do not account for the nonuniformity of the surface have only a slight effect on the classification accuracy; in other cases the classification accuracy descreases. The radiances above finite fields were computed to simulate radiances measured by a satellite. A simulation case including 11 agricultural fields and four natural fields (water, soil, savanah, and forest) was used to test the effect of the size of the background reflectance and the optical thickness of the atmosphere on classification accuracy. It is concluded that new atmospheric correction methods, which take into account the finite size of the fields, have to be developed to improve significantly the classification accuracy.

Rule-based land use/land cover classification in coastal areas using seasonal remote sensing imagery: a case study from Lianyungang City, China.

PubMed

Yang, Xiaoyan; Chen, Longgao; Li, Yingkui; Xi, Wenjia; Chen, Longqian

2015-07-01

Land use/land cover (LULC) inventory provides an important dataset in regional planning and environmental assessment. To efficiently obtain the LULC inventory, we compared the LULC classifications based on single satellite imagery with a rule-based classification based on multi-seasonal imagery in Lianyungang City, a coastal city in China, using CBERS-02 (the 2nd China-Brazil Environmental Resource Satellites) images. The overall accuracies of the classification based on single imagery are 78.9, 82.8, and 82.0% in winter, early summer, and autumn, respectively. The rule-based classification improves the accuracy to 87.9% (kappa 0.85), suggesting that combining multi-seasonal images can considerably improve the classification accuracy over any single image-based classification. This method could also be used to analyze seasonal changes of LULC types, especially for those associated with tidal changes in coastal areas. The distribution and inventory of LULC types with an overall accuracy of 87.9% and a spatial resolution of 19.5 m can assist regional planning and environmental assessment efficiently in Lianyungang City. This rule-based classification provides a guidance to improve accuracy for coastal areas with distinct LULC temporal spectral features.

Extending a field-based Sonoran desert vegetation classification to a regional scale using optical and microwave satellite imagery

NASA Astrophysics Data System (ADS)

Shupe, Scott Marshall

2000-10-01

Vegetation mapping in and regions facilitates ecological studies, land management, and provides a record to which future land changes can be compared. Accurate and representative mapping of desert vegetation requires a sound field sampling program and a methodology to transform the data collected into a representative classification system. Time and cost constraints require that a remote sensing approach be used if such a classification system is to be applied on a regional scale. However, desert vegetation may be sparse and thus difficult to sense at typical satellite resolutions, especially given the problem of soil reflectance. This study was designed to address these concerns by conducting vegetation mapping research using field and satellite data from the US Army Yuma Proving Ground (USYPG) in Southwest Arizona. Line and belt transect data from the Army's Land Condition Trend Analysis (LCTA) Program were transformed into relative cover and relative density classification schemes using cluster analysis. Ordination analysis of the same data produced two and three-dimensional graphs on which the homogeneity of each vegetation class could be examined. It was found that the use of correspondence analysis (CA), detrended correspondence analysis (DCA), and non-metric multidimensional scaling (NMS) ordination methods was superior to the use of any single ordination method for helping to clarify between-class and within-class relationships in vegetation composition. Analysis of these between-class and within-class relationships were of key importance in examining how well relative cover and relative density schemes characterize the USYPG vegetation. Using these two classification schemes as reference data, maximum likelihood and artificial neural net classifications were then performed on a coregistered dataset consisting of a summer Landsat Thematic Mapper (TM) image, one spring and one summer ERS-1 microwave image, and elevation, slope, and aspect layers. Classifications using a combination of ERS-1 imagery and elevation, slope, and aspect data were superior to classifications carried out using Landsat TM data alone. In all classification iterations it was consistently found that the highest classification accuracy was obtained by using a combination of Landsat TM, ERS-1, and elevation, slope, and aspect data. Maximum likelihood classification accuracy was found to be higher than artificial neural net classification in all cases.

Computer-Aided Diagnosis for Breast Ultrasound Using Computerized BI-RADS Features and Machine Learning Methods.

PubMed

Shan, Juan; Alam, S Kaisar; Garra, Brian; Zhang, Yingtao; Ahmed, Tahira

2016-04-01

This work identifies effective computable features from the Breast Imaging Reporting and Data System (BI-RADS), to develop a computer-aided diagnosis (CAD) system for breast ultrasound. Computerized features corresponding to ultrasound BI-RADs categories were designed and tested using a database of 283 pathology-proven benign and malignant lesions. Features were selected based on classification performance using a "bottom-up" approach for different machine learning methods, including decision tree, artificial neural network, random forest and support vector machine. Using 10-fold cross-validation on the database of 283 cases, the highest area under the receiver operating characteristic (ROC) curve (AUC) was 0.84 from a support vector machine with 77.7% overall accuracy; the highest overall accuracy, 78.5%, was from a random forest with the AUC 0.83. Lesion margin and orientation were optimum features common to all of the different machine learning methods. These features can be used in CAD systems to help distinguish benign from worrisome lesions. Copyright © 2016 World Federation for Ultrasound in Medicine & Biology. All rights reserved.

Visual spatial attention tracking using high-density SSVEP data for independent brain-computer communication.

PubMed

Kelly, Simon P; Lalor, Edmund C; Reilly, Richard B; Foxe, John J

2005-06-01

The steady-state visual evoked potential (SSVEP) has been employed successfully in brain-computer interface (BCI) research, but its use in a design entirely independent of eye movement has until recently not been reported. This paper presents strong evidence suggesting that the SSVEP can be used as an electrophysiological correlate of visual spatial attention that may be harnessed on its own or in conjunction with other correlates to achieve control in an independent BCI. In this study, 64-channel electroencephalography data were recorded from subjects who covertly attended to one of two bilateral flicker stimuli with superimposed letter sequences. Offline classification of left/right spatial attention was attempted by extracting SSVEPs at optimal channels selected for each subject on the basis of the scalp distribution of SSVEP magnitudes. This yielded an average accuracy of approximately 71% across ten subjects (highest 86%) comparable across two separate cases in which flicker frequencies were set within and outside the alpha range respectively. Further, combining SSVEP features with attention-dependent parieto-occipital alpha band modulations resulted in an average accuracy of 79% (highest 87%).

Interactive lesion segmentation with shape priors from offline and online learning.

PubMed

Shepherd, Tony; Prince, Simon J D; Alexander, Daniel C

2012-09-01

In medical image segmentation, tumors and other lesions demand the highest levels of accuracy but still call for the highest levels of manual delineation. One factor holding back automatic segmentation is the exemption of pathological regions from shape modelling techniques that rely on high-level shape information not offered by lesions. This paper introduces two new statistical shape models (SSMs) that combine radial shape parameterization with machine learning techniques from the field of nonlinear time series analysis. We then develop two dynamic contour models (DCMs) using the new SSMs as shape priors for tumor and lesion segmentation. From training data, the SSMs learn the lower level shape information of boundary fluctuations, which we prove to be nevertheless highly discriminant. One of the new DCMs also uses online learning to refine the shape prior for the lesion of interest based on user interactions. Classification experiments reveal superior sensitivity and specificity of the new shape priors over those previously used to constrain DCMs. User trials with the new interactive algorithms show that the shape priors are directly responsible for improvements in accuracy and reductions in user demand.

Boosted classification trees result in minor to modest improvement in the accuracy in classifying cardiovascular outcomes compared to conventional classification trees

PubMed Central

Austin, Peter C; Lee, Douglas S

2011-01-01

Purpose: Classification trees are increasingly being used to classifying patients according to the presence or absence of a disease or health outcome. A limitation of classification trees is their limited predictive accuracy. In the data-mining and machine learning literature, boosting has been developed to improve classification. Boosting with classification trees iteratively grows classification trees in a sequence of reweighted datasets. In a given iteration, subjects that were misclassified in the previous iteration are weighted more highly than subjects that were correctly classified. Classifications from each of the classification trees in the sequence are combined through a weighted majority vote to produce a final classification. The authors' objective was to examine whether boosting improved the accuracy of classification trees for predicting outcomes in cardiovascular patients. Methods: We examined the utility of boosting classification trees for classifying 30-day mortality outcomes in patients hospitalized with either acute myocardial infarction or congestive heart failure. Results: Improvements in the misclassification rate using boosted classification trees were at best minor compared to when conventional classification trees were used. Minor to modest improvements to sensitivity were observed, with only a negligible reduction in specificity. For predicting cardiovascular mortality, boosted classification trees had high specificity, but low sensitivity. Conclusions: Gains in predictive accuracy for predicting cardiovascular outcomes were less impressive than gains in performance observed in the data mining literature. PMID:22254181

Di-codon Usage for Gene Classification

NASA Astrophysics Data System (ADS)

Nguyen, Minh N.; Ma, Jianmin; Fogel, Gary B.; Rajapakse, Jagath C.

Classification of genes into biologically related groups facilitates inference of their functions. Codon usage bias has been described previously as a potential feature for gene classification. In this paper, we demonstrate that di-codon usage can further improve classification of genes. By using both codon and di-codon features, we achieve near perfect accuracies for the classification of HLA molecules into major classes and sub-classes. The method is illustrated on 1,841 HLA sequences which are classified into two major classes, HLA-I and HLA-II. Major classes are further classified into sub-groups. A binary SVM using di-codon usage patterns achieved 99.95% accuracy in the classification of HLA genes into major HLA classes; and multi-class SVM achieved accuracy rates of 99.82% and 99.03% for sub-class classification of HLA-I and HLA-II genes, respectively. Furthermore, by combining codon and di-codon usages, the prediction accuracies reached 100%, 99.82%, and 99.84% for HLA major class classification, and for sub-class classification of HLA-I and HLA-II genes, respectively.

Use of topographic and climatological models in a geographical data base to improve Landsat MSS classification for Olympic National Park

NASA Technical Reports Server (NTRS)

Cibula, William G.; Nyquist, Maurice O.

1987-01-01

An unsupervised computer classification of vegetation/landcover of Olympic National Park and surrounding environs was initially carried out using four bands of Landsat MSS data. The primary objective of the project was to derive a level of landcover classifications useful for park management applications while maintaining an acceptably high level of classification accuracy. Initially, nine generalized vegetation/landcover classes were derived. Overall classification accuracy was 91.7 percent. In an attempt to refine the level of classification, a geographic information system (GIS) approach was employed. Topographic data and watershed boundaries (inferred precipitation/temperature) data were registered with the Landsat MSS data. The resultant boolean operations yielded 21 vegetation/landcover classes while maintaining the same level of classification accuracy. The final classification provided much better identification and location of the major forest types within the park at the same high level of accuracy, and these met the project objective. This classification could now become inputs into a GIS system to help provide answers to park management coupled with other ancillary data programs such as fire management.

Efficient feature selection using a hybrid algorithm for the task of epileptic seizure detection

NASA Astrophysics Data System (ADS)

Lai, Kee Huong; Zainuddin, Zarita; Ong, Pauline

2014-07-01

Feature selection is a very important aspect in the field of machine learning. It entails the search of an optimal subset from a very large data set with high dimensional feature space. Apart from eliminating redundant features and reducing computational cost, a good selection of feature also leads to higher prediction and classification accuracy. In this paper, an efficient feature selection technique is introduced in the task of epileptic seizure detection. The raw data are electroencephalography (EEG) signals. Using discrete wavelet transform, the biomedical signals were decomposed into several sets of wavelet coefficients. To reduce the dimension of these wavelet coefficients, a feature selection method that combines the strength of both filter and wrapper methods is proposed. Principal component analysis (PCA) is used as part of the filter method. As for wrapper method, the evolutionary harmony search (HS) algorithm is employed. This metaheuristic method aims at finding the best discriminating set of features from the original data. The obtained features were then used as input for an automated classifier, namely wavelet neural networks (WNNs). The WNNs model was trained to perform a binary classification task, that is, to determine whether a given EEG signal was normal or epileptic. For comparison purposes, different sets of features were also used as input. Simulation results showed that the WNNs that used the features chosen by the hybrid algorithm achieved the highest overall classification accuracy.

Object based image analysis for the classification of the growth stages of Avocado crop, in Michoacán State, Mexico

NASA Astrophysics Data System (ADS)

Gao, Yan; Marpu, Prashanth; Morales Manila, Luis M.

2014-11-01

This paper assesses the suitability of 8-band Worldview-2 (WV2) satellite data and object-based random forest algorithm for the classification of avocado growth stages in Mexico. We tested both pixel-based with minimum distance (MD) and maximum likelihood (MLC) and object-based with Random Forest (RF) algorithm for this task. Training samples and verification data were selected by visual interpreting the WV2 images for seven thematic classes: fully grown, middle stage, and early stage of avocado crops, bare land, two types of natural forests, and water body. To examine the contribution of the four new spectral bands of WV2 sensor, all the tested classifications were carried out with and without the four new spectral bands. Classification accuracy assessment results show that object-based classification with RF algorithm obtained higher overall higher accuracy (93.06%) than pixel-based MD (69.37%) and MLC (64.03%) method. For both pixel-based and object-based methods, the classifications with the four new spectral bands (overall accuracy obtained higher accuracy than those without: overall accuracy of object-based RF classification with vs without: 93.06% vs 83.59%, pixel-based MD: 69.37% vs 67.2%, pixel-based MLC: 64.03% vs 36.05%, suggesting that the four new spectral bands in WV2 sensor contributed to the increase of the classification accuracy.

Accuracy assessment of seven global land cover datasets over China

NASA Astrophysics Data System (ADS)

Yang, Yongke; Xiao, Pengfeng; Feng, Xuezhi; Li, Haixing

2017-03-01

Land cover (LC) is the vital foundation to Earth science. Up to now, several global LC datasets have arisen with efforts of many scientific communities. To provide guidelines for data usage over China, nine LC maps from seven global LC datasets (IGBP DISCover, UMD, GLC, MCD12Q1, GLCNMO, CCI-LC, and GlobeLand30) were evaluated in this study. First, we compared their similarities and discrepancies in both area and spatial patterns, and analysed their inherent relations to data sources and classification schemes and methods. Next, five sets of validation sample units (VSUs) were collected to calculate their accuracy quantitatively. Further, we built a spatial analysis model and depicted their spatial variation in accuracy based on the five sets of VSUs. The results show that, there are evident discrepancies among these LC maps in both area and spatial patterns. For LC maps produced by different institutes, GLC 2000 and CCI-LC 2000 have the highest overall spatial agreement (53.8%). For LC maps produced by same institutes, overall spatial agreement of CCI-LC 2000 and 2010, and MCD12Q1 2001 and 2010 reach up to 99.8% and 73.2%, respectively; while more efforts are still needed if we hope to use these LC maps as time series data for model inputting, since both CCI-LC and MCD12Q1 fail to represent the rapid changing trend of several key LC classes in the early 21st century, in particular urban and built-up, snow and ice, water bodies, and permanent wetlands. With the highest spatial resolution, the overall accuracy of GlobeLand30 2010 is 82.39%. For the other six LC datasets with coarse resolution, CCI-LC 2010/2000 has the highest overall accuracy, and following are MCD12Q1 2010/2001, GLC 2000, GLCNMO 2008, IGBP DISCover, and UMD in turn. Beside that all maps exhibit high accuracy in homogeneous regions; local accuracies in other regions are quite different, particularly in Farming-Pastoral Zone of North China, mountains in Northeast China, and Southeast Hills. Special attention should be paid for data users who are interested in these regions.

A support vector machine classifier reduces interscanner variation in the HRCT classification of regional disease pattern in diffuse lung disease: Comparison to a Bayesian classifier

DOE Office of Scientific and Technical Information (OSTI.GOV)

Chang, Yongjun; Lim, Jonghyuck; Kim, Namkug

2013-05-15

Purpose: To investigate the effect of using different computed tomography (CT) scanners on the accuracy of high-resolution CT (HRCT) images in classifying regional disease patterns in patients with diffuse lung disease, support vector machine (SVM) and Bayesian classifiers were applied to multicenter data. Methods: Two experienced radiologists marked sets of 600 rectangular 20 Multiplication-Sign 20 pixel regions of interest (ROIs) on HRCT images obtained from two scanners (GE and Siemens), including 100 ROIs for each of local patterns of lungs-normal lung and five of regional pulmonary disease patterns (ground-glass opacity, reticular opacity, honeycombing, emphysema, and consolidation). Each ROI was assessedmore » using 22 quantitative features belonging to one of the following descriptors: histogram, gradient, run-length, gray level co-occurrence matrix, low-attenuation area cluster, and top-hat transform. For automatic classification, a Bayesian classifier and a SVM classifier were compared under three different conditions. First, classification accuracies were estimated using data from each scanner. Next, data from the GE and Siemens scanners were used for training and testing, respectively, and vice versa. Finally, all ROI data were integrated regardless of the scanner type and were then trained and tested together. All experiments were performed based on forward feature selection and fivefold cross-validation with 20 repetitions. Results: For each scanner, better classification accuracies were achieved with the SVM classifier than the Bayesian classifier (92% and 82%, respectively, for the GE scanner; and 92% and 86%, respectively, for the Siemens scanner). The classification accuracies were 82%/72% for training with GE data and testing with Siemens data, and 79%/72% for the reverse. The use of training and test data obtained from the HRCT images of different scanners lowered the classification accuracy compared to the use of HRCT images from the same scanner. For integrated ROI data obtained from both scanners, the classification accuracies with the SVM and Bayesian classifiers were 92% and 77%, respectively. The selected features resulting from the classification process differed by scanner, with more features included for the classification of the integrated HRCT data than for the classification of the HRCT data from each scanner. For the integrated data, consisting of HRCT images of both scanners, the classification accuracy based on the SVM was statistically similar to the accuracy of the data obtained from each scanner. However, the classification accuracy of the integrated data using the Bayesian classifier was significantly lower than the classification accuracy of the ROI data of each scanner. Conclusions: The use of an integrated dataset along with a SVM classifier rather than a Bayesian classifier has benefits in terms of the classification accuracy of HRCT images acquired with more than one scanner. This finding is of relevance in studies involving large number of images, as is the case in a multicenter trial with different scanners.« less

Corn and soybean Landsat MSS classification performance as a function of scene characteristics

NASA Technical Reports Server (NTRS)

Batista, G. T.; Hixson, M. M.; Bauer, M. E.

1982-01-01

In order to fully utilize remote sensing to inventory crop production, it is important to identify the factors that affect the accuracy of Landsat classifications. The objective of this study was to investigate the effect of scene characteristics involving crop, soil, and weather variables on the accuracy of Landsat classifications of corn and soybeans. Segments sampling the U.S. Corn Belt were classified using a Gaussian maximum likelihood classifier on multitemporally registered data from two key acquisition periods. Field size had a strong effect on classification accuracy with small fields tending to have low accuracies even when the effect of mixed pixels was eliminated. Other scene characteristics accounting for variability in classification accuracy included proportions of corn and soybeans, crop diversity index, proportion of all field crops, soil drainage, slope, soil order, long-term average soybean yield, maximum yield, relative position of the segment in the Corn Belt, weather, and crop development stage.

A Classification of Remote Sensing Image Based on Improved Compound Kernels of Svm

NASA Astrophysics Data System (ADS)

Zhao, Jianing; Gao, Wanlin; Liu, Zili; Mou, Guifen; Lu, Lin; Yu, Lina

The accuracy of RS classification based on SVM which is developed from statistical learning theory is high under small number of train samples, which results in satisfaction of classification on RS using SVM methods. The traditional RS classification method combines visual interpretation with computer classification. The accuracy of the RS classification, however, is improved a lot based on SVM method, because it saves much labor and time which is used to interpret images and collect training samples. Kernel functions play an important part in the SVM algorithm. It uses improved compound kernel function and therefore has a higher accuracy of classification on RS images. Moreover, compound kernel improves the generalization and learning ability of the kernel.

Nationwide forestry applications program. Analysis of forest classification accuracy

NASA Technical Reports Server (NTRS)

Congalton, R. G.; Mead, R. A.; Oderwald, R. G.; Heinen, J. (Principal Investigator)

1981-01-01

The development of LANDSAT classification accuracy assessment techniques, and of a computerized system for assessing wildlife habitat from land cover maps are considered. A literature review on accuracy assessment techniques and an explanation for the techniques development under both projects are included along with listings of the computer programs. The presentations and discussions at the National Working Conference on LANDSAT Classification Accuracy are summarized. Two symposium papers which were published on the results of this project are appended.

Assessment of Gait Characteristics in Total Knee Arthroplasty Patients Using a Hierarchical Partial Least Squares Method.

PubMed

Wang, Wei; Ackland, David C; McClelland, Jodie A; Webster, Kate E; Halgamuge, Saman

2018-01-01

Quantitative gait analysis is an important tool in objective assessment and management of total knee arthroplasty (TKA) patients. Studies evaluating gait patterns in TKA patients have tended to focus on discrete data such as spatiotemporal information, joint range of motion and peak values of kinematics and kinetics, or consider selected principal components of gait waveforms for analysis. These strategies may not have the capacity to capture small variations in gait patterns associated with each joint across an entire gait cycle, and may ultimately limit the accuracy of gait classification. The aim of this study was to develop an automatic feature extraction method to analyse patterns from high-dimensional autocorrelated gait waveforms. A general linear feature extraction framework was proposed and a hierarchical partial least squares method derived for discriminant analysis of multiple gait waveforms. The effectiveness of this strategy was verified using a dataset of joint angle and ground reaction force waveforms from 43 patients after TKA surgery and 31 healthy control subjects. Compared with principal component analysis and partial least squares methods, the hierarchical partial least squares method achieved generally better classification performance on all possible combinations of waveforms, with the highest classification accuracy . The novel hierarchical partial least squares method proposed is capable of capturing virtually all significant differences between TKA patients and the controls, and provides new insights into data visualization. The proposed framework presents a foundation for more rigorous classification of gait, and may ultimately be used to evaluate the effects of interventions such as surgery and rehabilitation.

Predicting temperate forest stand types using only structural profiles from discrete return airborne lidar

NASA Astrophysics Data System (ADS)

Fedrigo, Melissa; Newnham, Glenn J.; Coops, Nicholas C.; Culvenor, Darius S.; Bolton, Douglas K.; Nitschke, Craig R.

2018-02-01

Light detection and ranging (lidar) data have been increasingly used for forest classification due to its ability to penetrate the forest canopy and provide detail about the structure of the lower strata. In this study we demonstrate forest classification approaches using airborne lidar data as inputs to random forest and linear unmixing classification algorithms. Our results demonstrated that both random forest and linear unmixing models identified a distribution of rainforest and eucalypt stands that was comparable to existing ecological vegetation class (EVC) maps based primarily on manual interpretation of high resolution aerial imagery. Rainforest stands were also identified in the region that have not previously been identified in the EVC maps. The transition between stand types was better characterised by the random forest modelling approach. In contrast, the linear unmixing model placed greater emphasis on field plots selected as endmembers which may not have captured the variability in stand structure within a single stand type. The random forest model had the highest overall accuracy (84%) and Cohen's kappa coefficient (0.62). However, the classification accuracy was only marginally better than linear unmixing. The random forest model was applied to a region in the Central Highlands of south-eastern Australia to produce maps of stand type probability, including areas of transition (the 'ecotone') between rainforest and eucalypt forest. The resulting map provided a detailed delineation of forest classes, which specifically recognised the coalescing of stand types at the landscape scale. This represents a key step towards mapping the structural and spatial complexity of these ecosystems, which is important for both their management and conservation.

Effect of radiance-to-reflectance transformation and atmosphere removal on maximum likelihood classification accuracy of high-dimensional remote sensing data

NASA Technical Reports Server (NTRS)

Hoffbeck, Joseph P.; Landgrebe, David A.

1994-01-01

Many analysis algorithms for high-dimensional remote sensing data require that the remotely sensed radiance spectra be transformed to approximate reflectance to allow comparison with a library of laboratory reflectance spectra. In maximum likelihood classification, however, the remotely sensed spectra are compared to training samples, thus a transformation to reflectance may or may not be helpful. The effect of several radiance-to-reflectance transformations on maximum likelihood classification accuracy is investigated in this paper. We show that the empirical line approach, LOWTRAN7, flat-field correction, single spectrum method, and internal average reflectance are all non-singular affine transformations, and that non-singular affine transformations have no effect on discriminant analysis feature extraction and maximum likelihood classification accuracy. (An affine transformation is a linear transformation with an optional offset.) Since the Atmosphere Removal Program (ATREM) and the log residue method are not affine transformations, experiments with Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) data were conducted to determine the effect of these transformations on maximum likelihood classification accuracy. The average classification accuracy of the data transformed by ATREM and the log residue method was slightly less than the accuracy of the original radiance data. Since the radiance-to-reflectance transformations allow direct comparison of remotely sensed spectra with laboratory reflectance spectra, they can be quite useful in labeling the training samples required by maximum likelihood classification, but these transformations have only a slight effect or no effect at all on discriminant analysis and maximum likelihood classification accuracy.

Classifying Radio Galaxies with the Convolutional Neural Network

DOE Office of Scientific and Technical Information (OSTI.GOV)

Aniyan, A. K.; Thorat, K.

We present the application of a deep machine learning technique to classify radio images of extended sources on a morphological basis using convolutional neural networks (CNN). In this study, we have taken the case of the Fanaroff–Riley (FR) class of radio galaxies as well as radio galaxies with bent-tailed morphology. We have used archival data from the Very Large Array (VLA)—Faint Images of the Radio Sky at Twenty Centimeters survey and existing visually classified samples available in the literature to train a neural network for morphological classification of these categories of radio sources. Our training sample size for each of these categoriesmore » is ∼200 sources, which has been augmented by rotated versions of the same. Our study shows that CNNs can classify images of the FRI and FRII and bent-tailed radio galaxies with high accuracy (maximum precision at 95%) using well-defined samples and a “fusion classifier,” which combines the results of binary classifications, while allowing for a mechanism to find sources with unusual morphologies. The individual precision is highest for bent-tailed radio galaxies at 95% and is 91% and 75% for the FRI and FRII classes, respectively, whereas the recall is highest for FRI and FRIIs at 91% each, while the bent-tailed class has a recall of 79%. These results show that our results are comparable to that of manual classification, while being much faster. Finally, we discuss the computational and data-related challenges associated with the morphological classification of radio galaxies with CNNs.« less

Assessment of craniometric traits in South Indian dry skulls for sex determination.

PubMed

Ramamoorthy, Balakrishnan; Pai, Mangala M; Prabhu, Latha V; Muralimanju, B V; Rai, Rajalakshmi

2016-01-01

The skeleton plays an important role in sex determination in forensic anthropology. The skull bone is considered as the second best after the pelvic bone in sex determination due to its better retention of morphological features. Different populations have varying skeletal characteristics, making population specific analysis for sex determination essential. Hence the objective of this investigation is to obtain the accuracy of sex determination using cranial parameters of adult skulls to the highest percentage in South Indian population and to provide a baseline data for sex determination in South India. Seventy adult preserved human skulls were taken and based on the morphological traits were classified into 43 male skulls and 27 female skulls. A total of 26 craniometric parameters were studied. The data were analyzed by using the SPSS discriminant function. The analysis of stepwise, multivariate, and univariate discriminant function gave an accuracy of 77.1%, 85.7%, and 72.9% respectively. Multivariate direct discriminant function analysis classified skull bones into male and female with highest levels of accuracy. Using stepwise discriminant function analysis, the most dimorphic variable to determine sex of the skull, was biauricular breadth followed by weight. Subjecting the best dimorphic variables to univariate discriminant analysis, high levels of accuracy of sexual dimorphism was obtained. Percentage classification of high accuracies were obtained in this study indicating high level of sexual dimorphism in the crania, setting specific discriminant equations for the gender determination in South Indian people. Copyright © 2015 Elsevier Ltd and Faculty of Forensic and Legal Medicine. All rights reserved.

A fuzzy hill-climbing algorithm for the development of a compact associative classifier

NASA Astrophysics Data System (ADS)

Mitra, Soumyaroop; Lam, Sarah S.

2012-02-01

Classification, a data mining technique, has widespread applications including medical diagnosis, targeted marketing, and others. Knowledge discovery from databases in the form of association rules is one of the important data mining tasks. An integrated approach, classification based on association rules, has drawn the attention of the data mining community over the last decade. While attention has been mainly focused on increasing classifier accuracies, not much efforts have been devoted towards building interpretable and less complex models. This paper discusses the development of a compact associative classification model using a hill-climbing approach and fuzzy sets. The proposed methodology builds the rule-base by selecting rules which contribute towards increasing training accuracy, thus balancing classification accuracy with the number of classification association rules. The results indicated that the proposed associative classification model can achieve competitive accuracies on benchmark datasets with continuous attributes and lend better interpretability, when compared with other rule-based systems.

Comparison of ANN and SVM for classification of eye movements in EOG signals

NASA Astrophysics Data System (ADS)

Qi, Lim Jia; Alias, Norma

2018-03-01

Nowadays, electrooculogram is regarded as one of the most important biomedical signal in measuring and analyzing eye movement patterns. Thus, it is helpful in designing EOG-based Human Computer Interface (HCI). In this research, electrooculography (EOG) data was obtained from five volunteers. The (EOG) data was then preprocessed before feature extraction methods were employed to further reduce the dimensionality of data. Three feature extraction approaches were put forward, namely statistical parameters, autoregressive (AR) coefficients using Burg method, and power spectral density (PSD) using Yule-Walker method. These features would then become input to both artificial neural network (ANN) and support vector machine (SVM). The performance of the combination of different feature extraction methods and classifiers was presented and analyzed. It was found that statistical parameters + SVM achieved the highest classification accuracy of 69.75%.

Calibration and validation of the relative differenced Normalized Burn Ratio (RdNBR) to three measures of fire severity in the Sierra Nevada and Klamath Mountains, California, USA

USGS Publications Warehouse

Miller, J.D.; Knapp, E.E.; Key, C.H.; Skinner, C.N.; Isbell, C.J.; Creasy, R.M.; Sherlock, J.W.

2009-01-01

Multispectral satellite data have become a common tool used in the mapping of wildland fire effects. Fire severity, defined as the degree to which a site has been altered, is often the variable mapped. The Normalized Burn Ratio (NBR) used in an absolute difference change detection protocol (dNBR), has become the remote sensing method of choice for US Federal land management agencies to map fire severity due to wildland fire. However, absolute differenced vegetation indices are correlated to the pre-fire chlorophyll content of the vegetation occurring within the fire perimeter. Normalizing dNBR to produce a relativized dNBR (RdNBR) removes the biasing effect of the pre-fire condition. Employing RdNBR hypothetically allows creating categorical classifications using the same thresholds for fires occurring in similar vegetation types without acquiring additional calibration field data on each fire. In this paper we tested this hypothesis by developing thresholds on random training datasets, and then comparing accuracies for (1) fires that occurred within the same geographic region as the training dataset and in similar vegetation, and (2) fires from a different geographic region that is climatically and floristically similar to the training dataset region but supports more complex vegetation structure. We additionally compared map accuracies for three measures of fire severity: the composite burn index (CBI), percent change in tree canopy cover, and percent change in tree basal area. User's and producer's accuracies were highest for the most severe categories, ranging from 70.7% to 89.1%. Accuracies of the moderate fire severity category for measures describing effects only to trees (percent change in canopy cover and basal area) indicated that the classifications were generally not much better than random. Accuracies of the moderate category for the CBI classifications were somewhat better, averaging in the 50%-60% range. These results underscore the difficulty in isolating fire effects to individual vegetation strata when fire effects are mixed. We conclude that the models presented here and in Miller and Thode ([Miller, J.D. & Thode, A.E., (2007). Quantifying burn severity in a heterogeneous landscape with a relative version of the delta Normalized Burn Ratio (dNBR). Remote Sensing of Environment, 109, 66-80.]) can produce fire severity classifications (using either CBI, or percent change in canopy cover or basal area) that are of similar accuracy in fires not used in the original calibration process, at least in conifer dominated vegetation types in Mediterranean-climate California.

Derivation of an artificial gene to improve classification accuracy upon gene selection.

PubMed

Seo, Minseok; Oh, Sejong

2012-02-01

Classification analysis has been developed continuously since 1936. This research field has advanced as a result of development of classifiers such as KNN, ANN, and SVM, as well as through data preprocessing areas. Feature (gene) selection is required for very high dimensional data such as microarray before classification work. The goal of feature selection is to choose a subset of informative features that reduces processing time and provides higher classification accuracy. In this study, we devised a method of artificial gene making (AGM) for microarray data to improve classification accuracy. Our artificial gene was derived from a whole microarray dataset, and combined with a result of gene selection for classification analysis. We experimentally confirmed a clear improvement of classification accuracy after inserting artificial gene. Our artificial gene worked well for popular feature (gene) selection algorithms and classifiers. The proposed approach can be applied to any type of high dimensional dataset. Copyright © 2011 Elsevier Ltd. All rights reserved.

SVM-RFE based feature selection and Taguchi parameters optimization for multiclass SVM classifier.

PubMed

Huang, Mei-Ling; Hung, Yung-Hsiang; Lee, W M; Li, R K; Jiang, Bo-Ru

2014-01-01

Recently, support vector machine (SVM) has excellent performance on classification and prediction and is widely used on disease diagnosis or medical assistance. However, SVM only functions well on two-group classification problems. This study combines feature selection and SVM recursive feature elimination (SVM-RFE) to investigate the classification accuracy of multiclass problems for Dermatology and Zoo databases. Dermatology dataset contains 33 feature variables, 1 class variable, and 366 testing instances; and the Zoo dataset contains 16 feature variables, 1 class variable, and 101 testing instances. The feature variables in the two datasets were sorted in descending order by explanatory power, and different feature sets were selected by SVM-RFE to explore classification accuracy. Meanwhile, Taguchi method was jointly combined with SVM classifier in order to optimize parameters C and γ to increase classification accuracy for multiclass classification. The experimental results show that the classification accuracy can be more than 95% after SVM-RFE feature selection and Taguchi parameter optimization for Dermatology and Zoo databases.

SVM-RFE Based Feature Selection and Taguchi Parameters Optimization for Multiclass SVM Classifier

PubMed Central

Huang, Mei-Ling; Hung, Yung-Hsiang; Lee, W. M.; Li, R. K.; Jiang, Bo-Ru

2014-01-01

Recently, support vector machine (SVM) has excellent performance on classification and prediction and is widely used on disease diagnosis or medical assistance. However, SVM only functions well on two-group classification problems. This study combines feature selection and SVM recursive feature elimination (SVM-RFE) to investigate the classification accuracy of multiclass problems for Dermatology and Zoo databases. Dermatology dataset contains 33 feature variables, 1 class variable, and 366 testing instances; and the Zoo dataset contains 16 feature variables, 1 class variable, and 101 testing instances. The feature variables in the two datasets were sorted in descending order by explanatory power, and different feature sets were selected by SVM-RFE to explore classification accuracy. Meanwhile, Taguchi method was jointly combined with SVM classifier in order to optimize parameters C and γ to increase classification accuracy for multiclass classification. The experimental results show that the classification accuracy can be more than 95% after SVM-RFE feature selection and Taguchi parameter optimization for Dermatology and Zoo databases. PMID:25295306

SU-F-R-14: PET Based Radiomics to Predict Outcomes in Patients with Hodgkin Lymphoma

DOE Office of Scientific and Technical Information (OSTI.GOV)

Lee, J; Aristophanous, M; Akhtari, M

Purpose: To identify PET-based radiomics features associated with high refractory/relapsed disease risk for Hodgkin lymphoma patients. Methods: A total of 251 Hodgkin lymphoma patients including 19 primary refractory and 9 relapsed patients were investigated. All patients underwent an initial pre-treatment diagnostic FDG PET/CT scan. All cancerous lymph node regions (ROIs) were delineated by an experienced physician based on thresholding each volume of disease in the anatomical regions to SUV>2.5. We extracted 122 image features and evaluated the effect of ROI selection (the largest ROI, the ROI with highest mean SUV, merged ROI, and a single anatomic region [e.g. mediastinum]) onmore » classification accuracy. Random forest was used as a classifier and ROC analysis was used to assess the relationship between selected features and patient’s outcome status. Results: Each patient had between 1 and 9 separate ROIs, with much intra-patient variability in PET features. The best model, which used features from a single anatomic region (the mediastinal ROI, only volumes>5cc: 169 patients with 12 primary refractory) had a classification accuracy of 80.5% for primary refractory disease. The top five features, based on Gini index, consist of shape features (max 3D-diameter and volume) and texture features (correlation and information measure of correlation1&2). In the ROC analysis, sensitivity and specificity of the best model were 0.92 and 0.80, respectively. The area under the ROC (AUC) and the accuracy were 0.86 and 0.86, respectively. The classification accuracy was less than 60% for other ROI models or when ROIs less than 5cc were included. Conclusion: This study showed that PET-based radiomics features from the mediastinal lymph region are associated with primary refractory disease and therefore may play an important role in predicting outcomes in Hodgkin lymphoma patients. These features could be additive beyond baseline tumor and clinical characteristics, and may warrant more aggressive treatment.« less

Finding minimum gene subsets with heuristic breadth-first search algorithm for robust tumor classification

PubMed Central

2012-01-01

Background Previous studies on tumor classification based on gene expression profiles suggest that gene selection plays a key role in improving the classification performance. Moreover, finding important tumor-related genes with the highest accuracy is a very important task because these genes might serve as tumor biomarkers, which is of great benefit to not only tumor molecular diagnosis but also drug development. Results This paper proposes a novel gene selection method with rich biomedical meaning based on Heuristic Breadth-first Search Algorithm (HBSA) to find as many optimal gene subsets as possible. Due to the curse of dimensionality, this type of method could suffer from over-fitting and selection bias problems. To address these potential problems, a HBSA-based ensemble classifier is constructed using majority voting strategy from individual classifiers constructed by the selected gene subsets, and a novel HBSA-based gene ranking method is designed to find important tumor-related genes by measuring the significance of genes using their occurrence frequencies in the selected gene subsets. The experimental results on nine tumor datasets including three pairs of cross-platform datasets indicate that the proposed method can not only obtain better generalization performance but also find many important tumor-related genes. Conclusions It is found that the frequencies of the selected genes follow a power-law distribution, indicating that only a few top-ranked genes can be used as potential diagnosis biomarkers. Moreover, the top-ranked genes leading to very high prediction accuracy are closely related to specific tumor subtype and even hub genes. Compared with other related methods, the proposed method can achieve higher prediction accuracy with fewer genes. Moreover, they are further justified by analyzing the top-ranked genes in the context of individual gene function, biological pathway, and protein-protein interaction network. PMID:22830977

Enhancing the performance of regional land cover mapping

NASA Astrophysics Data System (ADS)

Wu, Weicheng; Zucca, Claudio; Karam, Fadi; Liu, Guangping

2016-10-01

Different pixel-based, object-based and subpixel-based methods such as time-series analysis, decision-tree, and different supervised approaches have been proposed to conduct land use/cover classification. However, despite their proven advantages in small dataset tests, their performance is variable and less satisfactory while dealing with large datasets, particularly, for regional-scale mapping with high resolution data due to the complexity and diversity in landscapes and land cover patterns, and the unacceptably long processing time. The objective of this paper is to demonstrate the comparatively highest performance of an operational approach based on integration of multisource information ensuring high mapping accuracy in large areas with acceptable processing time. The information used includes phenologically contrasted multiseasonal and multispectral bands, vegetation index, land surface temperature, and topographic features. The performance of different conventional and machine learning classifiers namely Malahanobis Distance (MD), Maximum Likelihood (ML), Artificial Neural Networks (ANNs), Support Vector Machines (SVMs) and Random Forests (RFs) was compared using the same datasets in the same IDL (Interactive Data Language) environment. An Eastern Mediterranean area with complex landscape and steep climate gradients was selected to test and develop the operational approach. The results showed that SVMs and RFs classifiers produced most accurate mapping at local-scale (up to 96.85% in Overall Accuracy), but were very time-consuming in whole-scene classification (more than five days per scene) whereas ML fulfilled the task rapidly (about 10 min per scene) with satisfying accuracy (94.2-96.4%). Thus, the approach composed of integration of seasonally contrasted multisource data and sampling at subclass level followed by a ML classification is a suitable candidate to become an operational and effective regional land cover mapping method.

Real-time, resource-constrained object classification on a micro-air vehicle

NASA Astrophysics Data System (ADS)

Buck, Louis; Ray, Laura

2013-12-01

A real-time embedded object classification algorithm is developed through the novel combination of binary feature descriptors, a bag-of-visual-words object model and the cortico-striatal loop (CSL) learning algorithm. The BRIEF, ORB and FREAK binary descriptors are tested and compared to SIFT descriptors with regard to their respective classification accuracies, execution times, and memory requirements when used with CSL on a 12.6 g ARM Cortex embedded processor running at 800 MHz. Additionally, the effect of x2 feature mapping and opponent-color representations used with these descriptors is examined. These tests are performed on four data sets of varying sizes and difficulty, and the BRIEF descriptor is found to yield the best combination of speed and classification accuracy. Its use with CSL achieves accuracies between 67% and 95% of those achieved with SIFT descriptors and allows for the embedded classification of a 128x192 pixel image in 0.15 seconds, 60 times faster than classification with SIFT. X2 mapping is found to provide substantial improvements in classification accuracy for all of the descriptors at little cost, while opponent-color descriptors are offer accuracy improvements only on colorful datasets.

A Nonparametric Approach to Estimate Classification Accuracy and Consistency

ERIC Educational Resources Information Center

Lathrop, Quinn N.; Cheng, Ying

2014-01-01

When cut scores for classifications occur on the total score scale, popular methods for estimating classification accuracy (CA) and classification consistency (CC) require assumptions about a parametric form of the test scores or about a parametric response model, such as item response theory (IRT). This article develops an approach to estimate CA…

Per-field crop classification in irrigated agricultural regions in middle Asia using random forest and support vector machine ensemble

NASA Astrophysics Data System (ADS)

Löw, Fabian; Schorcht, Gunther; Michel, Ulrich; Dech, Stefan; Conrad, Christopher

2012-10-01

Accurate crop identification and crop area estimation are important for studies on irrigated agricultural systems, yield and water demand modeling, and agrarian policy development. In this study a novel combination of Random Forest (RF) and Support Vector Machine (SVM) classifiers is presented that (i) enhances crop classification accuracy and (ii) provides spatial information on map uncertainty. The methodology was implemented over four distinct irrigated sites in Middle Asia using RapidEye time series data. The RF feature importance statistics was used as feature-selection strategy for the SVM to assess possible negative effects on classification accuracy caused by an oversized feature space. The results of the individual RF and SVM classifications were combined with rules based on posterior classification probability and estimates of classification probability entropy. SVM classification performance was increased by feature selection through RF. Further experimental results indicate that the hybrid classifier improves overall classification accuracy in comparison to the single classifiers as well as useŕs and produceŕs accuracy.

Automatic ICD-10 multi-class classification of cause of death from plaintext autopsy reports through expert-driven feature selection.

PubMed

Mujtaba, Ghulam; Shuib, Liyana; Raj, Ram Gopal; Rajandram, Retnagowri; Shaikh, Khairunisa; Al-Garadi, Mohammed Ali

2017-01-01

Widespread implementation of electronic databases has improved the accessibility of plaintext clinical information for supplementary use. Numerous machine learning techniques, such as supervised machine learning approaches or ontology-based approaches, have been employed to obtain useful information from plaintext clinical data. This study proposes an automatic multi-class classification system to predict accident-related causes of death from plaintext autopsy reports through expert-driven feature selection with supervised automatic text classification decision models. Accident-related autopsy reports were obtained from one of the largest hospital in Kuala Lumpur. These reports belong to nine different accident-related causes of death. Master feature vector was prepared by extracting features from the collected autopsy reports by using unigram with lexical categorization. This master feature vector was used to detect cause of death [according to internal classification of disease version 10 (ICD-10) classification system] through five automated feature selection schemes, proposed expert-driven approach, five subset sizes of features, and five machine learning classifiers. Model performance was evaluated using precisionM, recallM, F-measureM, accuracy, and area under ROC curve. Four baselines were used to compare the results with the proposed system. Random forest and J48 decision models parameterized using expert-driven feature selection yielded the highest evaluation measure approaching (85% to 90%) for most metrics by using a feature subset size of 30. The proposed system also showed approximately 14% to 16% improvement in the overall accuracy compared with the existing techniques and four baselines. The proposed system is feasible and practical to use for automatic classification of ICD-10-related cause of death from autopsy reports. The proposed system assists pathologists to accurately and rapidly determine underlying cause of death based on autopsy findings. Furthermore, the proposed expert-driven feature selection approach and the findings are generally applicable to other kinds of plaintext clinical reports.

Automatic ICD-10 multi-class classification of cause of death from plaintext autopsy reports through expert-driven feature selection

PubMed Central

Mujtaba, Ghulam; Shuib, Liyana; Raj, Ram Gopal; Rajandram, Retnagowri; Shaikh, Khairunisa; Al-Garadi, Mohammed Ali

2017-01-01

Objectives Widespread implementation of electronic databases has improved the accessibility of plaintext clinical information for supplementary use. Numerous machine learning techniques, such as supervised machine learning approaches or ontology-based approaches, have been employed to obtain useful information from plaintext clinical data. This study proposes an automatic multi-class classification system to predict accident-related causes of death from plaintext autopsy reports through expert-driven feature selection with supervised automatic text classification decision models. Methods Accident-related autopsy reports were obtained from one of the largest hospital in Kuala Lumpur. These reports belong to nine different accident-related causes of death. Master feature vector was prepared by extracting features from the collected autopsy reports by using unigram with lexical categorization. This master feature vector was used to detect cause of death [according to internal classification of disease version 10 (ICD-10) classification system] through five automated feature selection schemes, proposed expert-driven approach, five subset sizes of features, and five machine learning classifiers. Model performance was evaluated using precisionM, recallM, F-measureM, accuracy, and area under ROC curve. Four baselines were used to compare the results with the proposed system. Results Random forest and J48 decision models parameterized using expert-driven feature selection yielded the highest evaluation measure approaching (85% to 90%) for most metrics by using a feature subset size of 30. The proposed system also showed approximately 14% to 16% improvement in the overall accuracy compared with the existing techniques and four baselines. Conclusion The proposed system is feasible and practical to use for automatic classification of ICD-10-related cause of death from autopsy reports. The proposed system assists pathologists to accurately and rapidly determine underlying cause of death based on autopsy findings. Furthermore, the proposed expert-driven feature selection approach and the findings are generally applicable to other kinds of plaintext clinical reports. PMID:28166263

Information extraction with object based support vector machines and vegetation indices

NASA Astrophysics Data System (ADS)

Ustuner, Mustafa; Abdikan, Saygin; Balik Sanli, Fusun

2016-07-01

Information extraction through remote sensing data is important for policy and decision makers as extracted information provide base layers for many application of real world. Classification of remotely sensed data is the one of the most common methods of extracting information however it is still a challenging issue because several factors are affecting the accuracy of the classification. Resolution of the imagery, number and homogeneity of land cover classes, purity of training data and characteristic of adopted classifiers are just some of these challenging factors. Object based image classification has some superiority than pixel based classification for high resolution images since it uses geometry and structure information besides spectral information. Vegetation indices are also commonly used for the classification process since it provides additional spectral information for vegetation, forestry and agricultural areas. In this study, the impacts of the Normalized Difference Vegetation Index (NDVI) and Normalized Difference Red Edge Index (NDRE) on the classification accuracy of RapidEye imagery were investigated. Object based Support Vector Machines were implemented for the classification of crop types for the study area located in Aegean region of Turkey. Results demonstrated that the incorporation of NDRE increase the classification accuracy from 79,96% to 86,80% as overall accuracy, however NDVI decrease the classification accuracy from 79,96% to 78,90%. Moreover it is proven than object based classification with RapidEye data give promising results for crop type mapping and analysis.

Development of machine learning models for diagnosis of glaucoma.

PubMed

Kim, Seong Jae; Cho, Kyong Jin; Oh, Sejong

2017-01-01

The study aimed to develop machine learning models that have strong prediction power and interpretability for diagnosis of glaucoma based on retinal nerve fiber layer (RNFL) thickness and visual field (VF). We collected various candidate features from the examination of retinal nerve fiber layer (RNFL) thickness and visual field (VF). We also developed synthesized features from original features. We then selected the best features proper for classification (diagnosis) through feature evaluation. We used 100 cases of data as a test dataset and 399 cases of data as a training and validation dataset. To develop the glaucoma prediction model, we considered four machine learning algorithms: C5.0, random forest (RF), support vector machine (SVM), and k-nearest neighbor (KNN). We repeatedly composed a learning model using the training dataset and evaluated it by using the validation dataset. Finally, we got the best learning model that produces the highest validation accuracy. We analyzed quality of the models using several measures. The random forest model shows best performance and C5.0, SVM, and KNN models show similar accuracy. In the random forest model, the classification accuracy is 0.98, sensitivity is 0.983, specificity is 0.975, and AUC is 0.979. The developed prediction models show high accuracy, sensitivity, specificity, and AUC in classifying among glaucoma and healthy eyes. It will be used for predicting glaucoma against unknown examination records. Clinicians may reference the prediction results and be able to make better decisions. We may combine multiple learning models to increase prediction accuracy. The C5.0 model includes decision rules for prediction. It can be used to explain the reasons for specific predictions.

Improving the accuracy of k-nearest neighbor using local mean based and distance weight

NASA Astrophysics Data System (ADS)

Syaliman, K. U.; Nababan, E. B.; Sitompul, O. S.

2018-03-01

In k-nearest neighbor (kNN), the determination of classes for new data is normally performed by a simple majority vote system, which may ignore the similarities among data, as well as allowing the occurrence of a double majority class that can lead to misclassification. In this research, we propose an approach to resolve the majority vote issues by calculating the distance weight using a combination of local mean based k-nearest neighbor (LMKNN) and distance weight k-nearest neighbor (DWKNN). The accuracy of results is compared to the accuracy acquired from the original k-NN method using several datasets from the UCI Machine Learning repository, Kaggle and Keel, such as ionosphare, iris, voice genre, lower back pain, and thyroid. In addition, the proposed method is also tested using real data from a public senior high school in city of Tualang, Indonesia. Results shows that the combination of LMKNN and DWKNN was able to increase the classification accuracy of kNN, whereby the average accuracy on test data is 2.45% with the highest increase in accuracy of 3.71% occurring on the lower back pain symptoms dataset. For the real data, the increase in accuracy is obtained as high as 5.16%.

Comparing Pixel- and Object-Based Approaches in Effectively Classifying Wetland-Dominated Landscapes

PubMed Central

Berhane, Tedros M.; Lane, Charles R.; Wu, Qiusheng; Anenkhonov, Oleg A.; Chepinoga, Victor V.; Autrey, Bradley C.; Liu, Hongxing

2018-01-01

Wetland ecosystems straddle both terrestrial and aquatic habitats, performing many ecological functions directly and indirectly benefitting humans. However, global wetland losses are substantial. Satellite remote sensing and classification informs wise wetland management and monitoring. Both pixel- and object-based classification approaches using parametric and non-parametric algorithms may be effectively used in describing wetland structure and habitat, but which approach should one select? We conducted both pixel- and object-based image analyses (OBIA) using parametric (Iterative Self-Organizing Data Analysis Technique, ISODATA, and maximum likelihood, ML) and non-parametric (random forest, RF) approaches in the Barguzin Valley, a large wetland (~500 km2) in the Lake Baikal, Russia, drainage basin. Four Quickbird multispectral bands plus various spatial and spectral metrics (e.g., texture, Non-Differentiated Vegetation Index, slope, aspect, etc.) were analyzed using field-based regions of interest sampled to characterize an initial 18 ISODATA-based classes. Parsimoniously using a three-layer stack (Quickbird band 3, water ratio index (WRI), and mean texture) in the analyses resulted in the highest accuracy, 87.9% with pixel-based RF, followed by OBIA RF (segmentation scale 5, 84.6% overall accuracy), followed by pixel-based ML (83.9% overall accuracy). Increasing the predictors from three to five by adding Quickbird bands 2 and 4 decreased the pixel-based overall accuracy while increasing the OBIA RF accuracy to 90.4%. However, McNemar’s chi-square test confirmed no statistically significant difference in overall accuracy among the classifiers (pixel-based ML, RF, or object-based RF) for either the three- or five-layer analyses. Although potentially useful in some circumstances, the OBIA approach requires substantial resources and user input (such as segmentation scale selection—which was found to substantially affect overall accuracy). Hence, we conclude that pixel-based RF approaches are likely satisfactory for classifying wetland-dominated landscapes. PMID:29707381

Comparing Pixel- and Object-Based Approaches in Effectively Classifying Wetland-Dominated Landscapes.

PubMed

Berhane, Tedros M; Lane, Charles R; Wu, Qiusheng; Anenkhonov, Oleg A; Chepinoga, Victor V; Autrey, Bradley C; Liu, Hongxing

2018-01-01

Wetland ecosystems straddle both terrestrial and aquatic habitats, performing many ecological functions directly and indirectly benefitting humans. However, global wetland losses are substantial. Satellite remote sensing and classification informs wise wetland management and monitoring. Both pixel- and object-based classification approaches using parametric and non-parametric algorithms may be effectively used in describing wetland structure and habitat, but which approach should one select? We conducted both pixel- and object-based image analyses (OBIA) using parametric (Iterative Self-Organizing Data Analysis Technique, ISODATA, and maximum likelihood, ML) and non-parametric (random forest, RF) approaches in the Barguzin Valley, a large wetland (~500 km 2 ) in the Lake Baikal, Russia, drainage basin. Four Quickbird multispectral bands plus various spatial and spectral metrics (e.g., texture, Non-Differentiated Vegetation Index, slope, aspect, etc.) were analyzed using field-based regions of interest sampled to characterize an initial 18 ISODATA-based classes. Parsimoniously using a three-layer stack (Quickbird band 3, water ratio index (WRI), and mean texture) in the analyses resulted in the highest accuracy, 87.9% with pixel-based RF, followed by OBIA RF (segmentation scale 5, 84.6% overall accuracy), followed by pixel-based ML (83.9% overall accuracy). Increasing the predictors from three to five by adding Quickbird bands 2 and 4 decreased the pixel-based overall accuracy while increasing the OBIA RF accuracy to 90.4%. However, McNemar's chi-square test confirmed no statistically significant difference in overall accuracy among the classifiers (pixel-based ML, RF, or object-based RF) for either the three- or five-layer analyses. Although potentially useful in some circumstances, the OBIA approach requires substantial resources and user input (such as segmentation scale selection-which was found to substantially affect overall accuracy). Hence, we conclude that pixel-based RF approaches are likely satisfactory for classifying wetland-dominated landscapes.

A Pairwise Naïve Bayes Approach to Bayesian Classification.

PubMed

Asafu-Adjei, Josephine K; Betensky, Rebecca A

2015-10-01

Despite the relatively high accuracy of the naïve Bayes (NB) classifier, there may be several instances where it is not optimal, i.e. does not have the same classification performance as the Bayes classifier utilizing the joint distribution of the examined attributes. However, the Bayes classifier can be computationally intractable due to its required knowledge of the joint distribution. Therefore, we introduce a "pairwise naïve" Bayes (PNB) classifier that incorporates all pairwise relationships among the examined attributes, but does not require specification of the joint distribution. In this paper, we first describe the necessary and sufficient conditions under which the PNB classifier is optimal. We then discuss sufficient conditions for which the PNB classifier, and not NB, is optimal for normal attributes. Through simulation and actual studies, we evaluate the performance of our proposed classifier relative to the Bayes and NB classifiers, along with the HNB, AODE, LBR and TAN classifiers, using normal density and empirical estimation methods. Our applications show that the PNB classifier using normal density estimation yields the highest accuracy for data sets containing continuous attributes. We conclude that it offers a useful compromise between the Bayes and NB classifiers.

A comparative study of breast cancer diagnosis based on neural network ensemble via improved training algorithms.

PubMed

Azami, Hamed; Escudero, Javier

2015-08-01

Breast cancer is one of the most common types of cancer in women all over the world. Early diagnosis of this kind of cancer can significantly increase the chances of long-term survival. Since diagnosis of breast cancer is a complex problem, neural network (NN) approaches have been used as a promising solution. Considering the low speed of the back-propagation (BP) algorithm to train a feed-forward NN, we consider a number of improved NN trainings for the Wisconsin breast cancer dataset: BP with momentum, BP with adaptive learning rate, BP with adaptive learning rate and momentum, Polak-Ribikre conjugate gradient algorithm (CGA), Fletcher-Reeves CGA, Powell-Beale CGA, scaled CGA, resilient BP (RBP), one-step secant and quasi-Newton methods. An NN ensemble, which is a learning paradigm to combine a number of NN outputs, is used to improve the accuracy of the classification task. Results demonstrate that NN ensemble-based classification methods have better performance than NN-based algorithms. The highest overall average accuracy is 97.68% obtained by NN ensemble trained by RBP for 50%-50% training-test evaluation method.

Quantification of Reflection Patterns in Ground-Penetrating Radar Data

NASA Astrophysics Data System (ADS)

Moysey, S.; Knight, R. J.; Jol, H. M.; Allen-King, R. M.; Gaylord, D. R.

2005-12-01

Radar facies analysis provides a way of interpreting the large-scale structure of the subsurface from ground-penetrating radar (GPR) data. Radar facies are often distinguished from each other by the presence of patterns, such as flat-lying, dipping, or chaotic reflections, in different regions of a radar image. When these patterns can be associated with radar facies in a repeated and predictable manner we refer to them as `radar textures'. While it is often possible to qualitatively differentiate between radar textures visually, pattern recognition tools, like neural networks, require a quantitative measure to discriminate between them. We investigate whether currently available tools, such as instantaneous attributes or metrics adapted from standard texture analysis techniques, can be used to improve the classification of radar facies. To this end, we use a neural network to perform cross-validation tests that assess the efficacy of different textural measures for classifying radar facies in GPR data collected from the William River delta, Saskatchewan, Canada. We found that the highest classification accuracies (>93%) were obtained for measures of texture that preserve information about the spatial arrangement of reflections in the radar image, e.g., spatial covariance. Lower accuracy (87%) was obtained for classifications based directly on windows of amplitude data extracted from the radar image. Measures that did not account for the spatial arrangement of reflections in the image, e.g., instantaneous attributes and amplitude variance, yielded classification accuracies of less than 65%. Optimal classifications were obtained for textural measures that extracted sufficient information from the radar data to discriminate between radar facies but were insensitive to other facies specific characteristics. For example, the rotationally invariant Fourier-Mellin transform delivered better classification results than the spatial covariance because dip angle of the reflections, but not dip direction, was an important discriminator between radar facies at the William River delta. To extend the use of radar texture beyond the identification of radar facies to sedimentary facies we are investigating how sedimentary features are encoded in GPR data at Borden, Ontario, Canada. At this site, we have collected extensive sedimentary and hydrologic data over the area imaged by GPR. Analysis of this data coupled with synthetic modeling of the radar signal has allowed us to develop insight into the generation of radar texture in complex geologic environments.

Minimum distance classification in remote sensing

NASA Technical Reports Server (NTRS)

Wacker, A. G.; Landgrebe, D. A.

1972-01-01

The utilization of minimum distance classification methods in remote sensing problems, such as crop species identification, is considered. Literature concerning both minimum distance classification problems and distance measures is reviewed. Experimental results are presented for several examples. The objective of these examples is to: (a) compare the sample classification accuracy of a minimum distance classifier, with the vector classification accuracy of a maximum likelihood classifier, and (b) compare the accuracy of a parametric minimum distance classifier with that of a nonparametric one. Results show the minimum distance classifier performance is 5% to 10% better than that of the maximum likelihood classifier. The nonparametric classifier is only slightly better than the parametric version.

Seasonal discrimination of C3 and C4 grasses functional types: An evaluation of the prospects of varying spectral configurations of new generation sensors

NASA Astrophysics Data System (ADS)

Shoko, Cletah; Mutanga, Onisimo

2017-10-01

The present study assessed the potential of varying spectral configuration of Landsat 8 Operational Land Imager (OLI), Sentinel 2 MultiSpectal Instrument (MSI) and Worldview 2 sensors in the seasonal discrimination of Festuca costata (C3) and Themeda Triandra (C4) grass species in the Drakensberg, South Africa. This was achieved by resampling hyperspectral measurements to the spectral windows corresponding to the three sensors at two distinct seasonal periods (summer peak and end of winter), using the Discriminant Analysis (DA) classification ensemble. In summer, standard bands of the Worldview 2 produced the highest overall classification accuracy (98.61%), followed by Sentinel 2 (97.52%), whereas the Landsat 8 spectral configuration was the least performer, using vegetation indices (95.83%). In winter, Sentinel 2 spectral bands produced the highest accuracy (96.18%) for the two species, followed by Worldview 2 (94.44%) and Landsat 8 yielded the least (91.67%) accuracy. Results also showed that maximum separability between C3 and C4 grasses was in summer, while at the end of winter considerable overlaps were noted, especially when using the spectral settings of the Landsat 8 OLI and Sentinel 2 shortwave infrared bands. Test of significance in species reflectance further confirmed that in summer, there were significant differences (P < 0.05), whereas in winter, most of the spectral windows of all sensors yielded insignificant differences (P > 0.05) between the two species. In this regard, the peak summer period presents a promising opportunity for the spectral discrimination of C3 and C4 grass species functional types, than the end of winter, when using multispectral sensors. Results from this study highlight the influence of seasonality on discrimination and therefore provide the basis for the successful discrimination and mapping of C3 and C4 grass species.

Detection of Periodic Leg Movements by Machine Learning Methods Using Polysomnographic Parameters Other Than Leg Electromyography

PubMed Central

Umut, İlhan; Çentik, Güven

2016-01-01

The number of channels used for polysomnographic recording frequently causes difficulties for patients because of the many cables connected. Also, it increases the risk of having troubles during recording process and increases the storage volume. In this study, it is intended to detect periodic leg movement (PLM) in sleep with the use of the channels except leg electromyography (EMG) by analysing polysomnography (PSG) data with digital signal processing (DSP) and machine learning methods. PSG records of 153 patients of different ages and genders with PLM disorder diagnosis were examined retrospectively. A novel software was developed for the analysis of PSG records. The software utilizes the machine learning algorithms, statistical methods, and DSP methods. In order to classify PLM, popular machine learning methods (multilayer perceptron, K-nearest neighbour, and random forests) and logistic regression were used. Comparison of classified results showed that while K-nearest neighbour classification algorithm had higher average classification rate (91.87%) and lower average classification error value (RMSE = 0.2850), multilayer perceptron algorithm had the lowest average classification rate (83.29%) and the highest average classification error value (RMSE = 0.3705). Results showed that PLM can be classified with high accuracy (91.87%) without leg EMG record being present. PMID:27213008

Detection of Periodic Leg Movements by Machine Learning Methods Using Polysomnographic Parameters Other Than Leg Electromyography.

PubMed

Umut, İlhan; Çentik, Güven

2016-01-01

The number of channels used for polysomnographic recording frequently causes difficulties for patients because of the many cables connected. Also, it increases the risk of having troubles during recording process and increases the storage volume. In this study, it is intended to detect periodic leg movement (PLM) in sleep with the use of the channels except leg electromyography (EMG) by analysing polysomnography (PSG) data with digital signal processing (DSP) and machine learning methods. PSG records of 153 patients of different ages and genders with PLM disorder diagnosis were examined retrospectively. A novel software was developed for the analysis of PSG records. The software utilizes the machine learning algorithms, statistical methods, and DSP methods. In order to classify PLM, popular machine learning methods (multilayer perceptron, K-nearest neighbour, and random forests) and logistic regression were used. Comparison of classified results showed that while K-nearest neighbour classification algorithm had higher average classification rate (91.87%) and lower average classification error value (RMSE = 0.2850), multilayer perceptron algorithm had the lowest average classification rate (83.29%) and the highest average classification error value (RMSE = 0.3705). Results showed that PLM can be classified with high accuracy (91.87%) without leg EMG record being present.

A Ternary Hybrid EEG-NIRS Brain-Computer Interface for the Classification of Brain Activation Patterns during Mental Arithmetic, Motor Imagery, and Idle State.

PubMed

Shin, Jaeyoung; Kwon, Jinuk; Im, Chang-Hwan

2018-01-01

The performance of a brain-computer interface (BCI) can be enhanced by simultaneously using two or more modalities to record brain activity, which is generally referred to as a hybrid BCI. To date, many BCI researchers have tried to implement a hybrid BCI system by combining electroencephalography (EEG) and functional near-infrared spectroscopy (NIRS) to improve the overall accuracy of binary classification. However, since hybrid EEG-NIRS BCI, which will be denoted by hBCI in this paper, has not been applied to ternary classification problems, paradigms and classification strategies appropriate for ternary classification using hBCI are not well investigated. Here we propose the use of an hBCI for the classification of three brain activation patterns elicited by mental arithmetic, motor imagery, and idle state, with the aim to elevate the information transfer rate (ITR) of hBCI by increasing the number of classes while minimizing the loss of accuracy. EEG electrodes were placed over the prefrontal cortex and the central cortex, and NIRS optodes were placed only on the forehead. The ternary classification problem was decomposed into three binary classification problems using the "one-versus-one" (OVO) classification strategy to apply the filter-bank common spatial patterns filter to EEG data. A 10 × 10-fold cross validation was performed using shrinkage linear discriminant analysis (sLDA) to evaluate the average classification accuracies for EEG-BCI, NIRS-BCI, and hBCI when the meta-classification method was adopted to enhance classification accuracy. The ternary classification accuracies for EEG-BCI, NIRS-BCI, and hBCI were 76.1 ± 12.8, 64.1 ± 9.7, and 82.2 ± 10.2%, respectively. The classification accuracy of the proposed hBCI was thus significantly higher than those of the other BCIs ( p < 0.005). The average ITR for the proposed hBCI was calculated to be 4.70 ± 1.92 bits/minute, which was 34.3% higher than that reported for a previous binary hBCI study.

Selective classification for improved robustness of myoelectric control under nonideal conditions.

PubMed

Scheme, Erik J; Englehart, Kevin B; Hudgins, Bernard S

2011-06-01

Recent literature in pattern recognition-based myoelectric control has highlighted a disparity between classification accuracy and the usability of upper limb prostheses. This paper suggests that the conventionally defined classification accuracy may be idealistic and may not reflect true clinical performance. Herein, a novel myoelectric control system based on a selective multiclass one-versus-one classification scheme, capable of rejecting unknown data patterns, is introduced. This scheme is shown to outperform nine other popular classifiers when compared using conventional classification accuracy as well as a form of leave-one-out analysis that may be more representative of real prosthetic use. Additionally, the classification scheme allows for real-time, independent adjustment of individual class-pair boundaries making it flexible and intuitive for clinical use.

Multi-source remotely sensed data fusion for improving land cover classification

NASA Astrophysics Data System (ADS)

Chen, Bin; Huang, Bo; Xu, Bing

2017-02-01

Although many advances have been made in past decades, land cover classification of fine-resolution remotely sensed (RS) data integrating multiple temporal, angular, and spectral features remains limited, and the contribution of different RS features to land cover classification accuracy remains uncertain. We proposed to improve land cover classification accuracy by integrating multi-source RS features through data fusion. We further investigated the effect of different RS features on classification performance. The results of fusing Landsat-8 Operational Land Imager (OLI) data with Moderate Resolution Imaging Spectroradiometer (MODIS), China Environment 1A series (HJ-1A), and Advanced Spaceborne Thermal Emission and Reflection (ASTER) digital elevation model (DEM) data, showed that the fused data integrating temporal, spectral, angular, and topographic features achieved better land cover classification accuracy than the original RS data. Compared with the topographic feature, the temporal and angular features extracted from the fused data played more important roles in classification performance, especially those temporal features containing abundant vegetation growth information, which markedly increased the overall classification accuracy. In addition, the multispectral and hyperspectral fusion successfully discriminated detailed forest types. Our study provides a straightforward strategy for hierarchical land cover classification by making full use of available RS data. All of these methods and findings could be useful for land cover classification at both regional and global scales.

Applying support vector machine on hybrid fNIRS/EEG signal to classify driver's conditions (Conference Presentation)

NASA Astrophysics Data System (ADS)

Nguyen, Thien; Ahn, Sangtae; Jang, Hyojung; Jun, Sung C.; Kim, Jae G.

2016-03-01

Driver's condition plays a critical role in driving safety. The fact that about 20 percent of automobile accidents occurred due to driver fatigue leads to a demand for developing a method to monitor driver's status. In this study, we acquired brain signals such as oxy- and deoxyhemoglobin and neuronal electrical activity by a hybrid fNIRS/EEG system. Experiments were conducted with 11 subjects under two conditions: Normal condition, when subjects had enough sleep, and sleep deprivation condition, when subject did not sleep previous night. During experiment, subject performed a driving task with a car simulation system for 30 minutes. After experiment, oxy-hemoglobin and deoxy-hemoglobin changes were derived from fNIRS data, while beta and alpha band relative power were calculated from EEG data. Decrement of oxy-hemoglobin, beta band power, and increment of alpha band power were found in sleep deprivation condition compare to normal condition. These features were then applied to classify two conditions by Fisher's linear discriminant analysis (FLDA). The ratio of alpha-beta relative power showed classification accuracy with a range between 62% and 99% depending on a subject. However, utilization of both EEG and fNIRS features increased accuracy in the range between 68% and 100%. The highest increase of accuracy is from 63% using EEG to 99% using both EEG and fNIRS features. In conclusion, the enhancement of classification accuracy is shown by adding a feature from fNIRS to the feature from EEG using FLDA which provides the need of developing a hybrid fNIRS/EEG system.

Real-data comparison of data mining methods in prediction of diabetes in iran.

PubMed

Tapak, Lily; Mahjub, Hossein; Hamidi, Omid; Poorolajal, Jalal

2013-09-01

Diabetes is one of the most common non-communicable diseases in developing countries. Early screening and diagnosis play an important role in effective prevention strategies. This study compared two traditional classification methods (logistic regression and Fisher linear discriminant analysis) and four machine-learning classifiers (neural networks, support vector machines, fuzzy c-mean, and random forests) to classify persons with and without diabetes. The data set used in this study included 6,500 subjects from the Iranian national non-communicable diseases risk factors surveillance obtained through a cross-sectional survey. The obtained sample was based on cluster sampling of the Iran population which was conducted in 2005-2009 to assess the prevalence of major non-communicable disease risk factors. Ten risk factors that are commonly associated with diabetes were selected to compare the performance of six classifiers in terms of sensitivity, specificity, total accuracy, and area under the receiver operating characteristic (ROC) curve criteria. Support vector machines showed the highest total accuracy (0.986) as well as area under the ROC (0.979). Also, this method showed high specificity (1.000) and sensitivity (0.820). All other methods produced total accuracy of more than 85%, but for all methods, the sensitivity values were very low (less than 0.350). The results of this study indicate that, in terms of sensitivity, specificity, and overall classification accuracy, the support vector machine model ranks first among all the classifiers tested in the prediction of diabetes. Therefore, this approach is a promising classifier for predicting diabetes, and it should be further investigated for the prediction of other diseases.

Vulnerable land ecosystems classification using spatial context and spectral indices

NASA Astrophysics Data System (ADS)

Ibarrola-Ulzurrun, Edurne; Gonzalo-Martín, Consuelo; Marcello, Javier

2017-10-01

Natural habitats are exposed to growing pressure due to intensification of land use and tourism development. Thus, obtaining information on the vegetation is necessary for conservation and management projects. In this context, remote sensing is an important tool for monitoring and managing habitats, being classification a crucial stage. The majority of image classifications techniques are based upon the pixel-based approach. An alternative is the object-based (OBIA) approach, in which a previous segmentation step merges image pixels to create objects that are then classified. Besides, improved results may be gained by incorporating additional spatial information and specific spectral indices into the classification process. The main goal of this work was to implement and assess object-based classification techniques on very-high resolution imagery incorporating spectral indices and contextual spatial information in the classification models. The study area was Teide National Park in Canary Islands (Spain) using Worldview-2 orthoready imagery. In the classification model, two common indices were selected Normalized Difference Vegetation Index (NDVI) and Optimized Soil Adjusted Vegetation Index (OSAVI), as well as two specific Worldview-2 sensor indices, Worldview Vegetation Index and Worldview Soil Index. To include the contextual information, Grey Level Co-occurrence Matrices (GLCM) were used. The classification was performed training a Support Vector Machine with sufficient and representative number of vegetation samples (Spartocytisus supranubius, Pterocephalus lasiospermus, Descurainia bourgaeana and Pinus canariensis) as well as urban, road and bare soil classes. Confusion Matrices were computed to evaluate the results from each classification model obtaining the highest overall accuracy (90.07%) combining both Worldview indices with the GLCM-dissimilarity.

Automated classification of tropical shrub species: a hybrid of leaf shape and machine learning approach

PubMed Central

Murat, Miraemiliana; Abu, Arpah; Yap, Hwa Jen; Yong, Kien-Thai

2017-01-01

Plants play a crucial role in foodstuff, medicine, industry, and environmental protection. The skill of recognising plants is very important in some applications, including conservation of endangered species and rehabilitation of lands after mining activities. However, it is a difficult task to identify plant species because it requires specialized knowledge. Developing an automated classification system for plant species is necessary and valuable since it can help specialists as well as the public in identifying plant species easily. Shape descriptors were applied on the myDAUN dataset that contains 45 tropical shrub species collected from the University of Malaya (UM), Malaysia. Based on literature review, this is the first study in the development of tropical shrub species image dataset and classification using a hybrid of leaf shape and machine learning approach. Four types of shape descriptors were used in this study namely morphological shape descriptors (MSD), Histogram of Oriented Gradients (HOG), Hu invariant moments (Hu) and Zernike moments (ZM). Single descriptor, as well as the combination of hybrid descriptors were tested and compared. The tropical shrub species are classified using six different classifiers, which are artificial neural network (ANN), random forest (RF), support vector machine (SVM), k-nearest neighbour (k-NN), linear discriminant analysis (LDA) and directed acyclic graph multiclass least squares twin support vector machine (DAG MLSTSVM). In addition, three types of feature selection methods were tested in the myDAUN dataset, Relief, Correlation-based feature selection (CFS) and Pearson’s coefficient correlation (PCC). The well-known Flavia dataset and Swedish Leaf dataset were used as the validation dataset on the proposed methods. The results showed that the hybrid of all descriptors of ANN outperformed the other classifiers with an average classification accuracy of 98.23% for the myDAUN dataset, 95.25% for the Flavia dataset and 99.89% for the Swedish Leaf dataset. In addition, the Relief feature selection method achieved the highest classification accuracy of 98.13% after 80 (or 60%) of the original features were reduced, from 133 to 53 descriptors in the myDAUN dataset with the reduction in computational time. Subsequently, the hybridisation of four descriptors gave the best results compared to others. It is proven that the combination MSD and HOG were good enough for tropical shrubs species classification. Hu and ZM descriptors also improved the accuracy in tropical shrubs species classification in terms of invariant to translation, rotation and scale. ANN outperformed the others for tropical shrub species classification in this study. Feature selection methods can be used in the classification of tropical shrub species, as the comparable results could be obtained with the reduced descriptors and reduced in computational time and cost. PMID:28924506

Automated classification of tropical shrub species: a hybrid of leaf shape and machine learning approach.

PubMed

Murat, Miraemiliana; Chang, Siow-Wee; Abu, Arpah; Yap, Hwa Jen; Yong, Kien-Thai

2017-01-01

Plants play a crucial role in foodstuff, medicine, industry, and environmental protection. The skill of recognising plants is very important in some applications, including conservation of endangered species and rehabilitation of lands after mining activities. However, it is a difficult task to identify plant species because it requires specialized knowledge. Developing an automated classification system for plant species is necessary and valuable since it can help specialists as well as the public in identifying plant species easily. Shape descriptors were applied on the myDAUN dataset that contains 45 tropical shrub species collected from the University of Malaya (UM), Malaysia. Based on literature review, this is the first study in the development of tropical shrub species image dataset and classification using a hybrid of leaf shape and machine learning approach. Four types of shape descriptors were used in this study namely morphological shape descriptors (MSD), Histogram of Oriented Gradients (HOG), Hu invariant moments (Hu) and Zernike moments (ZM). Single descriptor, as well as the combination of hybrid descriptors were tested and compared. The tropical shrub species are classified using six different classifiers, which are artificial neural network (ANN), random forest (RF), support vector machine (SVM), k-nearest neighbour (k-NN), linear discriminant analysis (LDA) and directed acyclic graph multiclass least squares twin support vector machine (DAG MLSTSVM). In addition, three types of feature selection methods were tested in the myDAUN dataset, Relief, Correlation-based feature selection (CFS) and Pearson's coefficient correlation (PCC). The well-known Flavia dataset and Swedish Leaf dataset were used as the validation dataset on the proposed methods. The results showed that the hybrid of all descriptors of ANN outperformed the other classifiers with an average classification accuracy of 98.23% for the myDAUN dataset, 95.25% for the Flavia dataset and 99.89% for the Swedish Leaf dataset. In addition, the Relief feature selection method achieved the highest classification accuracy of 98.13% after 80 (or 60%) of the original features were reduced, from 133 to 53 descriptors in the myDAUN dataset with the reduction in computational time. Subsequently, the hybridisation of four descriptors gave the best results compared to others. It is proven that the combination MSD and HOG were good enough for tropical shrubs species classification. Hu and ZM descriptors also improved the accuracy in tropical shrubs species classification in terms of invariant to translation, rotation and scale. ANN outperformed the others for tropical shrub species classification in this study. Feature selection methods can be used in the classification of tropical shrub species, as the comparable results could be obtained with the reduced descriptors and reduced in computational time and cost.

Competitive Deep-Belief Networks for Underwater Acoustic Target Recognition

PubMed Central

Shen, Sheng; Yao, Xiaohui; Sheng, Meiping; Wang, Chen

2018-01-01

Underwater acoustic target recognition based on ship-radiated noise belongs to the small-sample-size recognition problems. A competitive deep-belief network is proposed to learn features with more discriminative information from labeled and unlabeled samples. The proposed model consists of four stages: (1) A standard restricted Boltzmann machine is pretrained using a large number of unlabeled data to initialize its parameters; (2) the hidden units are grouped according to categories, which provides an initial clustering model for competitive learning; (3) competitive training and back-propagation algorithms are used to update the parameters to accomplish the task of clustering; (4) by applying layer-wise training and supervised fine-tuning, a deep neural network is built to obtain features. Experimental results show that the proposed method can achieve classification accuracy of 90.89%, which is 8.95% higher than the accuracy obtained by the compared methods. In addition, the highest accuracy of our method is obtained with fewer features than other methods. PMID:29570642

Examining change detection approaches for tropical mangrove monitoring

USGS Publications Warehouse

Myint, Soe W.; Franklin, Janet; Buenemann, Michaela; Kim, Won; Giri, Chandra

2014-01-01

This study evaluated the effectiveness of different band combinations and classifiers (unsupervised, supervised, object-oriented nearest neighbor, and object-oriented decision rule) for quantifying mangrove forest change using multitemporal Landsat data. A discriminant analysis using spectra of different vegetation types determined that bands 2 (0.52 to 0.6 μm), 5 (1.55 to 1.75 μm), and 7 (2.08 to 2.35 μm) were the most effective bands for differentiating mangrove forests from surrounding land cover types. A ranking of thirty-six change maps, produced by comparing the classification accuracy of twelve change detection approaches, was used. The object-based Nearest Neighbor classifier produced the highest mean overall accuracy (84 percent) regardless of band combinations. The automated decision rule-based approach (mean overall accuracy of 88 percent) as well as a composite of bands 2, 5, and 7 used with the unsupervised classifier and the same composite or all band difference with the object-oriented Nearest Neighbor classifier were the most effective approaches.

Comparison of wheat classification accuracy using different classifiers of the image-100 system

NASA Technical Reports Server (NTRS)

Dejesusparada, N. (Principal Investigator); Chen, S. C.; Moreira, M. A.; Delima, A. M.

1981-01-01

Classification results using single-cell and multi-cell signature acquisition options, a point-by-point Gaussian maximum-likelihood classifier, and K-means clustering of the Image-100 system are presented. Conclusions reached are that: a better indication of correct classification can be provided by using a test area which contains various cover types of the study area; classification accuracy should be evaluated considering both the percentages of correct classification and error of commission; supervised classification approaches are better than K-means clustering; Gaussian distribution maximum likelihood classifier is better than Single-cell and Multi-cell Signature Acquisition Options of the Image-100 system; and in order to obtain a high classification accuracy in a large and heterogeneous crop area, using Gaussian maximum-likelihood classifier, homogeneous spectral subclasses of the study crop should be created to derive training statistics.

Impact of atmospheric correction and image filtering on hyperspectral classification of tree species using support vector machine

NASA Astrophysics Data System (ADS)

Shahriari Nia, Morteza; Wang, Daisy Zhe; Bohlman, Stephanie Ann; Gader, Paul; Graves, Sarah J.; Petrovic, Milenko

2015-01-01

Hyperspectral images can be used to identify savannah tree species at the landscape scale, which is a key step in measuring biomass and carbon, and tracking changes in species distributions, including invasive species, in these ecosystems. Before automated species mapping can be performed, image processing and atmospheric correction is often performed, which can potentially affect the performance of classification algorithms. We determine how three processing and correction techniques (atmospheric correction, Gaussian filters, and shade/green vegetation filters) affect the prediction accuracy of classification of tree species at pixel level from airborne visible/infrared imaging spectrometer imagery of longleaf pine savanna in Central Florida, United States. Species classification using fast line-of-sight atmospheric analysis of spectral hypercubes (FLAASH) atmospheric correction outperformed ATCOR in the majority of cases. Green vegetation (normalized difference vegetation index) and shade (near-infrared) filters did not increase classification accuracy when applied to large and continuous patches of specific species. Finally, applying a Gaussian filter reduces interband noise and increases species classification accuracy. Using the optimal preprocessing steps, our classification accuracy of six species classes is about 75%.

Comparison of Single and Multi-Scale Method for Leaf and Wood Points Classification from Terrestrial Laser Scanning Data

NASA Astrophysics Data System (ADS)

Wei, Hongqiang; Zhou, Guiyun; Zhou, Junjie

2018-04-01

The classification of leaf and wood points is an essential preprocessing step for extracting inventory measurements and canopy characterization of trees from the terrestrial laser scanning (TLS) data. The geometry-based approach is one of the widely used classification method. In the geometry-based method, it is common practice to extract salient features at one single scale before the features are used for classification. It remains unclear how different scale(s) used affect the classification accuracy and efficiency. To assess the scale effect on the classification accuracy and efficiency, we extracted the single-scale and multi-scale salient features from the point clouds of two oak trees of different sizes and conducted the classification on leaf and wood. Our experimental results show that the balanced accuracy of the multi-scale method is higher than the average balanced accuracy of the single-scale method by about 10 % for both trees. The average speed-up ratio of single scale classifiers over multi-scale classifier for each tree is higher than 30.

Comparing Features for Classification of MEG Responses to Motor Imagery.

PubMed

Halme, Hanna-Leena; Parkkonen, Lauri

2016-01-01

Motor imagery (MI) with real-time neurofeedback could be a viable approach, e.g., in rehabilitation of cerebral stroke. Magnetoencephalography (MEG) noninvasively measures electric brain activity at high temporal resolution and is well-suited for recording oscillatory brain signals. MI is known to modulate 10- and 20-Hz oscillations in the somatomotor system. In order to provide accurate feedback to the subject, the most relevant MI-related features should be extracted from MEG data. In this study, we evaluated several MEG signal features for discriminating between left- and right-hand MI and between MI and rest. MEG was measured from nine healthy participants imagining either left- or right-hand finger tapping according to visual cues. Data preprocessing, feature extraction and classification were performed offline. The evaluated MI-related features were power spectral density (PSD), Morlet wavelets, short-time Fourier transform (STFT), common spatial patterns (CSP), filter-bank common spatial patterns (FBCSP), spatio-spectral decomposition (SSD), and combined SSD+CSP, CSP+PSD, CSP+Morlet, and CSP+STFT. We also compared four classifiers applied to single trials using 5-fold cross-validation for evaluating the classification accuracy and its possible dependence on the classification algorithm. In addition, we estimated the inter-session left-vs-right accuracy for each subject. The SSD+CSP combination yielded the best accuracy in both left-vs-right (mean 73.7%) and MI-vs-rest (mean 81.3%) classification. CSP+Morlet yielded the best mean accuracy in inter-session left-vs-right classification (mean 69.1%). There were large inter-subject differences in classification accuracy, and the level of the 20-Hz suppression correlated significantly with the subjective MI-vs-rest accuracy. Selection of the classification algorithm had only a minor effect on the results. We obtained good accuracy in sensor-level decoding of MI from single-trial MEG data. Feature extraction methods utilizing both the spatial and spectral profile of MI-related signals provided the best classification results, suggesting good performance of these methods in an online MEG neurofeedback system.

PCA based feature reduction to improve the accuracy of decision tree c4.5 classification

NASA Astrophysics Data System (ADS)

Nasution, M. Z. F.; Sitompul, O. S.; Ramli, M.

2018-03-01

Splitting attribute is a major process in Decision Tree C4.5 classification. However, this process does not give a significant impact on the establishment of the decision tree in terms of removing irrelevant features. It is a major problem in decision tree classification process called over-fitting resulting from noisy data and irrelevant features. In turns, over-fitting creates misclassification and data imbalance. Many algorithms have been proposed to overcome misclassification and overfitting on classifications Decision Tree C4.5. Feature reduction is one of important issues in classification model which is intended to remove irrelevant data in order to improve accuracy. The feature reduction framework is used to simplify high dimensional data to low dimensional data with non-correlated attributes. In this research, we proposed a framework for selecting relevant and non-correlated feature subsets. We consider principal component analysis (PCA) for feature reduction to perform non-correlated feature selection and Decision Tree C4.5 algorithm for the classification. From the experiments conducted using available data sets from UCI Cervical cancer data set repository with 858 instances and 36 attributes, we evaluated the performance of our framework based on accuracy, specificity and precision. Experimental results show that our proposed framework is robust to enhance classification accuracy with 90.70% accuracy rates.

Comparison of Machine Learning Methods for the Arterial Hypertension Diagnostics

PubMed Central

Belo, David; Gamboa, Hugo

2017-01-01

The paper presents results of machine learning approach accuracy applied analysis of cardiac activity. The study evaluates the diagnostics possibilities of the arterial hypertension by means of the short-term heart rate variability signals. Two groups were studied: 30 relatively healthy volunteers and 40 patients suffering from the arterial hypertension of II-III degree. The following machine learning approaches were studied: linear and quadratic discriminant analysis, k-nearest neighbors, support vector machine with radial basis, decision trees, and naive Bayes classifier. Moreover, in the study, different methods of feature extraction are analyzed: statistical, spectral, wavelet, and multifractal. All in all, 53 features were investigated. Investigation results show that discriminant analysis achieves the highest classification accuracy. The suggested approach of noncorrelated feature set search achieved higher results than data set based on the principal components. PMID:28831239

Meat and Fish Freshness Inspection System Based on Odor Sensing

PubMed Central

Hasan, Najam ul; Ejaz, Naveed; Ejaz, Waleed; Kim, Hyung Seok

2012-01-01

We propose a method for building a simple electronic nose based on commercially available sensors used to sniff in the market and identify spoiled/contaminated meat stocked for sale in butcher shops. Using a metal oxide semiconductor-based electronic nose, we measured the smell signature from two of the most common meat foods (beef and fish) stored at room temperature. Food samples were divided into two groups: fresh beef with decayed fish and fresh fish with decayed beef. The prime objective was to identify the decayed item using the developed electronic nose. Additionally, we tested the electronic nose using three pattern classification algorithms (artificial neural network, support vector machine and k-nearest neighbor), and compared them based on accuracy, sensitivity, and specificity. The results demonstrate that the k-nearest neighbor algorithm has the highest accuracy. PMID:23202222

Improving crop classification through attention to the timing of airborne radar acquisitions

NASA Technical Reports Server (NTRS)

Brisco, B.; Ulaby, F. T.; Protz, R.

1984-01-01

Radar remote sensors may provide valuable input to crop classification procedures because of (1) their independence of weather conditions and solar illumination, and (2) their ability to respond to differences in crop type. Manual classification of multidate synthetic aperture radar (SAR) imagery resulted in an overall accuracy of 83 percent for corn, forest, grain, and 'other' cover types. Forests and corn fields were identified with accuracies approaching or exceeding 90 percent. Grain fields and 'other' fields were often confused with each other, resulting in classification accuracies of 51 and 66 percent, respectively. The 83 percent correct classification represents a 10 percent improvement when compared to similar SAR data for the same area collected at alternate time periods in 1978. These results demonstrate that improvements in crop classification accuracy can be achieved with SAR data by synchronizing data collection times with crop growth stages in order to maximize differences in the geometric and dielectric properties of the cover types of interest.

IMPACTS OF PATCH SIZE AND LANDSCAPE HETEROGENEITY ON THEMATIC IMAGE CLASSIFICATION ACCURACY

EPA Science Inventory

Impacts of Patch Size and Landscape Heterogeneity on Thematic Image Classification Accuracy.
Currently, most thematic accuracy assessments of classified remotely sensed images oily account for errors between the various classes employed, at particular pixels of interest, thu...

Development of a method for classification of hospitals based on results of the diagnosis-related groups and the principle of case-mix index.

PubMed

Shelf, Louay

2016-08-18

This research used DRGs and CMI to adjust medical waste production through the calculation of DRGs and CMI scores. These scores were used to assess the performances of teaching hospitals in Damascus. The linear correlations between these scores and the annual amount of waste and DRGs values were studied. The differences between the daily waste generations before and after the adjustment process were determined. Accordingly, the highest values of DRGs and CMI Scores were for the pediatric and Al Assad hospitals respectively. Among the teaching hospitals in Damascus, Al Assad has achieved the highest performance. Based on the results, the accuracy and homogeneity of medical waste generation rates were improved, which in turn leads to continuous improvement in the management of medical wastes.

Differentiation chronic post traumatic stress disorder patients from healthy subjects using objective and subjective sleep-related parameters.

PubMed

Tahmasian, Masoud; Jamalabadi, Hamidreza; Abedini, Mina; Ghadami, Mohammad R; Sepehry, Amir A; Knight, David C; Khazaie, Habibolah

2017-05-22

Sleep disturbance is common in chronic post-traumatic stress disorder (PTSD). However, prior work has demonstrated that there are inconsistencies between subjective and objective assessments of sleep disturbance in PTSD. Therefore, we investigated whether subjective or objective sleep assessment has greater clinical utility to differentiate PTSD patients from healthy subjects. Further, we evaluated whether the combination of subjective and objective methods improves the accuracy of classification into patient versus healthy groups, which has important diagnostic implications. We recruited 32 chronic war-induced PTSD patients and 32 age- and gender-matched healthy subjects to participate in this study. Subjective (i.e. from three self-reported sleep questionnaires) and objective sleep-related data (i.e. from actigraphy scores) were collected from each participant. Subjective, objective, and combined (subjective and objective) sleep data were then analyzed using support vector machine classification. The classification accuracy, sensitivity, and specificity for subjective variables were 89.2%, 89.3%, and 89%, respectively. The classification accuracy, sensitivity, and specificity for objective variables were 65%, 62.3%, and 67.8%, respectively. The classification accuracy, sensitivity, and specificity for the aggregate variables (combination of subjective and objective variables) were 91.6%, 93.0%, and 90.3%, respectively. Our findings indicate that classification accuracy using subjective measurements is superior to objective measurements and the combination of both assessments appears to improve the classification accuracy for differentiating PTSD patients from healthy individuals. Copyright © 2017 Elsevier B.V. All rights reserved.

Segmentation schema for enhancing land cover identification: A case study using Sentinel 2 data

NASA Astrophysics Data System (ADS)

Mongus, Domen; Žalik, Borut

2018-04-01

Land monitoring is performed increasingly using high and medium resolution optical satellites, such as the Sentinel-2. However, optical data is inevitably subjected to the variable operational conditions under which it was acquired. Overlapping of features caused by shadows, soft transitions between shadowed and non-shadowed regions, and temporal variability of the observed land-cover types require radiometric corrections. This study examines a new approach to enhancing the accuracy of land cover identification that resolves this problem. The proposed method constructs an ensemble-type classification model with weak classifiers tuned to the particular operational conditions under which the data was acquired. Iterative segmentation over the learning set is applied for this purpose, where feature space is partitioned according to the likelihood of misclassifications introduced by the classification model. As these are a consequence of overlapping features, such partitioning avoids the need for radiometric corrections of the data, and divides land cover types implicitly into subclasses. As a result, improved performance of all tested classification approaches were measured during the validation that was conducted on Sentinel-2 data. The highest accuracies in terms of F1-scores were achieved using the Naive Bayes Classifier as the weak classifier, while supplementing original spectral signatures with normalised difference vegetation index and texture analysis features, namely, average intensity, contrast, homogeneity, and dissimilarity. In total, an F1-score of nearly 95% was achieved in this way, with F1-scores of each particular land cover type reaching above 90%.

Investigation of different classifiers and channel configurations of a mobile P300-based brain-computer interface.

PubMed

Ludwig, Simone A; Kong, Jun

2017-12-01

Innovative methods and new technologies have significantly improved the quality of our daily life. However, disabled people, for example those that cannot use their arms and legs anymore, often cannot benefit from these developments, since they cannot use their hands to interact with traditional interaction methods (such as mouse or keyboard) to communicate with a computer system. A brain-computer interface (BCI) system allows such a disabled person to control an external device via brain waves. Past research mostly dealt with static interfaces, which limit users to a stationary location. However, since we are living in a world that is highly mobile, this paper evaluates a speller interface on a mobile phone used in a moving condition. The spelling experiments were conducted with 14 able-bodied subjects using visual flashes as the stimulus to spell 47 alphanumeric characters (38 letters and 9 numbers). This data was then used for the classification experiments. In par- ticular, two research directions are pursued. The first investigates the impact of different classification algorithms, and the second direction looks at the channel configuration, i.e., which channels are most beneficial in terms of achieving the highest classification accuracy. The evaluation results indicate that the Bayesian Linear Discriminant Analysis algorithm achieves the best accuracy. Also, the findings of the investigation on the channel configuration, which can potentially reduce the amount of data processing on a mobile device with limited computing capacity, is especially useful in mobile BCIs.

Can Preoperative Magnetic Resonance Imaging Predict the Reparability of Massive Rotator Cuff Tears?

PubMed

Kim, Jung Youn; Park, Ji Seon; Rhee, Yong Girl

2017-06-01

Numerous studies have shown preoperative fatty infiltration of rotator cuff muscles to be strongly negatively correlated with the successful repair of massive rotator cuff tears (RCTs). To assess the association between factors identified on preoperative magnetic resonance imaging (MRI), especially infraspinatus fatty infiltration, and the reparability of massive RCTs. Case-control study; Level of evidence, 3. We analyzed a total of 105 patients with massive RCTs for whom MRI was performed ≤6 months before arthroscopic procedures. The mean age of the patients was 62.7 years (range, 46-83 years), and 46 were men. Among them, complete repair was possible in 50 patients (48%) and not possible in 55 patients (52%). The tangent sign, fatty infiltration of the rotator cuff, and Patte classification were evaluated as predictors of reparability. Using the receiver operating characteristic curve and the area under the curve (AUC), the prediction accuracy of each variable and combinations of variables were measured. Reparability was associated with fatty infiltration of the supraspinatus ( P = .0045) and infraspinatus ( P < .001) muscles, the tangent sign ( P = .0033), and the Patte classification ( P < .001) but not with fatty infiltration of the subscapularis and teres minor ( P = .425 and .132, respectively). The cut-off values for supraspinatus and infraspinatus fatty infiltration were grade >3 and grade >2, respectively. The examination of single variables revealed that infraspinatus fatty infiltration showed the highest AUC value (0.812; sensitivity: 0.86; specificity: 0.76), while the tangent sign showed the lowest AUC value (0.626; sensitivity: 0.38; specificity: 0.87). Among 2-variable combinations, the combination of infraspinatus fatty infiltration and the Patte classification showed the highest AUC value (0.874; sensitivity: 0.54; specificity: 0.96). The combination of 4 variables, that is, infraspinatus and supraspinatus fatty infiltration, the tangent sign, and the Patte classification, had an AUC of 0.866 (sensitivity: 0.28; specificity: 0.98), which was lower than the highest AUC value (0.874; sensitivity: 0.54; specificity: 0.96) among the 2-variable combinations. The tangent sign or Patte classification alone was not a predictive indicator of the reparability of massive RCTs. Among single variables, infraspinatus fatty infiltration was the most effective in predicting reparability, while the combination of Goutallier classification <3 of the infraspinatus and Patte classification ≤2 of the rotator cuff muscles was the most predictive among the combinations of variables. This information may help predict the reparability of massive RCTs.

Analysis of thematic mapper simulator data collected over eastern North Dakota

NASA Technical Reports Server (NTRS)

Anderson, J. E. (Principal Investigator)

1982-01-01

The results of the analysis of aircraft-acquired thematic mapper simulator (TMS) data, collected to investigate the utility of thematic mapper data in crop area and land cover estimates, are discussed. Results of the analysis indicate that the seven-channel TMS data are capable of delineating the 13 crop types included in the study to an overall pixel classification accuracy of 80.97% correct, with relative efficiencies for four crop types examined between 1.62 and 26.61. Both supervised and unsupervised spectral signature development techniques were evaluated. The unsupervised methods proved to be inferior (based on analysis of variance) for the majority of crop types considered. Given the ground truth data set used for spectral signature development as well as evaluation of performance, it is possible to demonstrate which signature development technique would produce the highest percent correct classification for each crop type.

Assessment of Cropping System Diversity in the Fergana Valley Through Image Fusion of Landsat 8 and SENTINEL-1

NASA Astrophysics Data System (ADS)

Dimov, D.; Kuhn, J.; Conrad, C.

2016-06-01

In the transitioning agricultural societies of the world, food security is an essential element of livelihood and economic development with the agricultural sector very often being the major employment factor and income source. Rapid population growth, urbanization, pollution, desertification, soil degradation and climate change pose a variety of threats to a sustainable agricultural development and can be expressed as agricultural vulnerability components. Diverse cropping patterns may help to adapt the agricultural systems to those hazards in terms of increasing the potential yield and resilience to water scarcity. Thus, the quantification of crop diversity using indices like the Simpson Index of Diversity (SID) e.g. through freely available remote sensing data becomes a very important issue. This however requires accurate land use classifications. In this study, the focus is set on the cropping system diversity of garden plots, summer crop fields and orchard plots which are the prevalent agricultural systems in the test area of the Fergana Valley in Uzbekistan. In order to improve the accuracy of land use classification algorithms with low or medium resolution data, a novel processing chain through the hitherto unique fusion of optical and SAR data from the Landsat 8 and Sentinel-1 platforms is proposed. The combination of both sensors is intended to enhance the object's textural and spectral signature rather than just to enhance the spatial context through pansharpening. It could be concluded that the Ehlers fusion algorithm gave the most suitable results. Based on the derived image fusion different object-based image classification algorithms such as SVM, Naïve Bayesian and Random Forest were evaluated whereby the latter one achieved the highest classification accuracy. Subsequently, the SID was applied to measure the diversification of the three main cropping systems.

EEG-Based Brain-Computer Interface for Decoding Motor Imagery Tasks within the Same Hand Using Choi-Williams Time-Frequency Distribution

PubMed Central

Alwanni, Hisham; Baslan, Yara; Alnuman, Nasim; Daoud, Mohammad I.

2017-01-01

This paper presents an EEG-based brain-computer interface system for classifying eleven motor imagery (MI) tasks within the same hand. The proposed system utilizes the Choi-Williams time-frequency distribution (CWD) to construct a time-frequency representation (TFR) of the EEG signals. The constructed TFR is used to extract five categories of time-frequency features (TFFs). The TFFs are processed using a hierarchical classification model to identify the MI task encapsulated within the EEG signals. To evaluate the performance of the proposed approach, EEG data were recorded for eighteen intact subjects and four amputated subjects while imagining to perform each of the eleven hand MI tasks. Two performance evaluation analyses, namely channel- and TFF-based analyses, are conducted to identify the best subset of EEG channels and the TFFs category, respectively, that enable the highest classification accuracy between the MI tasks. In each evaluation analysis, the hierarchical classification model is trained using two training procedures, namely subject-dependent and subject-independent procedures. These two training procedures quantify the capability of the proposed approach to capture both intra- and inter-personal variations in the EEG signals for different MI tasks within the same hand. The results demonstrate the efficacy of the approach for classifying the MI tasks within the same hand. In particular, the classification accuracies obtained for the intact and amputated subjects are as high as 88.8% and 90.2%, respectively, for the subject-dependent training procedure, and 80.8% and 87.8%, respectively, for the subject-independent training procedure. These results suggest the feasibility of applying the proposed approach to control dexterous prosthetic hands, which can be of great benefit for individuals suffering from hand amputations. PMID:28832513

Influence of pansharpening techniques in obtaining accurate vegetation thematic maps

NASA Astrophysics Data System (ADS)

Ibarrola-Ulzurrun, Edurne; Gonzalo-Martin, Consuelo; Marcello-Ruiz, Javier

2016-10-01

In last decades, there have been a decline in natural resources, becoming important to develop reliable methodologies for their management. The appearance of very high resolution sensors has offered a practical and cost-effective means for a good environmental management. In this context, improvements are needed for obtaining higher quality of the information available in order to get reliable classified images. Thus, pansharpening enhances the spatial resolution of the multispectral band by incorporating information from the panchromatic image. The main goal in the study is to implement pixel and object-based classification techniques applied to the fused imagery using different pansharpening algorithms and the evaluation of thematic maps generated that serve to obtain accurate information for the conservation of natural resources. A vulnerable heterogenic ecosystem from Canary Islands (Spain) was chosen, Teide National Park, and Worldview-2 high resolution imagery was employed. The classes considered of interest were set by the National Park conservation managers. 7 pansharpening techniques (GS, FIHS, HCS, MTF based, Wavelet `à trous' and Weighted Wavelet `à trous' through Fractal Dimension Maps) were chosen in order to improve the data quality with the goal to analyze the vegetation classes. Next, different classification algorithms were applied at pixel-based and object-based approach, moreover, an accuracy assessment of the different thematic maps obtained were performed. The highest classification accuracy was obtained applying Support Vector Machine classifier at object-based approach in the Weighted Wavelet `à trous' through Fractal Dimension Maps fused image. Finally, highlight the difficulty of the classification in Teide ecosystem due to the heterogeneity and the small size of the species. Thus, it is important to obtain accurate thematic maps for further studies in the management and conservation of natural resources.

Global Optimization Ensemble Model for Classification Methods

PubMed Central

Anwar, Hina; Qamar, Usman; Muzaffar Qureshi, Abdul Wahab

2014-01-01

Supervised learning is the process of data mining for deducing rules from training datasets. A broad array of supervised learning algorithms exists, every one of them with its own advantages and drawbacks. There are some basic issues that affect the accuracy of classifier while solving a supervised learning problem, like bias-variance tradeoff, dimensionality of input space, and noise in the input data space. All these problems affect the accuracy of classifier and are the reason that there is no global optimal method for classification. There is not any generalized improvement method that can increase the accuracy of any classifier while addressing all the problems stated above. This paper proposes a global optimization ensemble model for classification methods (GMC) that can improve the overall accuracy for supervised learning problems. The experimental results on various public datasets showed that the proposed model improved the accuracy of the classification models from 1% to 30% depending upon the algorithm complexity. PMID:24883382

Investigating the Capability of IRS-P6-LISS IV Satellite Image for Pistachio Forests Density Mapping (case Study: Northeast of Iran)

NASA Astrophysics Data System (ADS)

Hoseini, F.; Darvishsefat, A. A.; Zargham, N.

2012-07-01

In order to investigate the capability of satellite images for Pistachio forests density mapping, IRS-P6-LISS IV data were analyzed in an area of 500 ha in Iran. After geometric correction, suitable training areas were determined based on fieldwork. Suitable spectral transformations like NDVI, PVI and PCA were performed. A ground truth map included of 34 plots (each plot 1 ha) were prepared. Hard and soft supervised classifications were performed with 5 density classes (0-5%, 5-10%, 10-15%, 15-20% and > 20%). Because of low separability of classes, some classes were merged and classifications were repeated with 3 classes. Finally, the highest overall accuracy and kappa coefficient of 70% and 0.44, respectively, were obtained with three classes (0-5%, 5-20%, and > 20%) by fuzzy classifier. Considering the low kappa value obtained, it could be concluded that the result of the classification was not desirable. Therefore, this approach is not appropriate for operational mapping of these valuable Pistachio forests.

From genus to phylum: large-subunit and internal transcribed spacer rRNA operon regions show similar classification accuracies influenced by database composition.

PubMed

Porras-Alfaro, Andrea; Liu, Kuan-Liang; Kuske, Cheryl R; Xie, Gary

2014-02-01

We compared the classification accuracy of two sections of the fungal internal transcribed spacer (ITS) region, individually and combined, and the 5' section (about 600 bp) of the large-subunit rRNA (LSU), using a naive Bayesian classifier and BLASTN. A hand-curated ITS-LSU training set of 1,091 sequences and a larger training set of 8,967 ITS region sequences were used. Of the factors evaluated, database composition and quality had the largest effect on classification accuracy, followed by fragment size and use of a bootstrap cutoff to improve classification confidence. The naive Bayesian classifier and BLASTN gave similar results at higher taxonomic levels, but the classifier was faster and more accurate at the genus level when a bootstrap cutoff was used. All of the ITS and LSU sections performed well (>97.7% accuracy) at higher taxonomic ranks from kingdom to family, and differences between them were small at the genus level (within 0.66 to 1.23%). When full-length sequence sections were used, the LSU outperformed the ITS1 and ITS2 fragments at the genus level, but the ITS1 and ITS2 showed higher accuracy when smaller fragment sizes of the same length and a 50% bootstrap cutoff were used. In a comparison using the larger ITS training set, ITS1 and ITS2 had very similar accuracy classification for fragments between 100 and 200 bp. Collectively, the results show that any of the ITS or LSU sections we tested provided comparable classification accuracy to the genus level and underscore the need for larger and more diverse classification training sets.

From Genus to Phylum: Large-Subunit and Internal Transcribed Spacer rRNA Operon Regions Show Similar Classification Accuracies Influenced by Database Composition

PubMed Central

Liu, Kuan-Liang; Kuske, Cheryl R.

2014-01-01

We compared the classification accuracy of two sections of the fungal internal transcribed spacer (ITS) region, individually and combined, and the 5′ section (about 600 bp) of the large-subunit rRNA (LSU), using a naive Bayesian classifier and BLASTN. A hand-curated ITS-LSU training set of 1,091 sequences and a larger training set of 8,967 ITS region sequences were used. Of the factors evaluated, database composition and quality had the largest effect on classification accuracy, followed by fragment size and use of a bootstrap cutoff to improve classification confidence. The naive Bayesian classifier and BLASTN gave similar results at higher taxonomic levels, but the classifier was faster and more accurate at the genus level when a bootstrap cutoff was used. All of the ITS and LSU sections performed well (>97.7% accuracy) at higher taxonomic ranks from kingdom to family, and differences between them were small at the genus level (within 0.66 to 1.23%). When full-length sequence sections were used, the LSU outperformed the ITS1 and ITS2 fragments at the genus level, but the ITS1 and ITS2 showed higher accuracy when smaller fragment sizes of the same length and a 50% bootstrap cutoff were used. In a comparison using the larger ITS training set, ITS1 and ITS2 had very similar accuracy classification for fragments between 100 and 200 bp. Collectively, the results show that any of the ITS or LSU sections we tested provided comparable classification accuracy to the genus level and underscore the need for larger and more diverse classification training sets. PMID:24242255

GHS additivity formula: can it predict the acute systemic toxicity of agrochemical formulations that contain acutely toxic ingredients?

PubMed

Van Cott, Andrew; Hastings, Charles E; Landsiedel, Robert; Kolle, Susanne; Stinchcombe, Stefan

2018-02-01

In vivo acute systemic testing is a regulatory requirement for agrochemical formulations. GHS specifies an alternative computational approach (GHS additivity formula) for calculating the acute toxicity of mixtures. We collected acute systemic toxicity data from formulations that contained one of several acutely-toxic active ingredients. The resulting acute data set includes 210 formulations tested for oral toxicity, 128 formulations tested for inhalation toxicity and 31 formulations tested for dermal toxicity. The GHS additivity formula was applied to each of these formulations and compared with the experimental in vivo result. In the acute oral assay, the GHS additivity formula misclassified 110 formulations using the GHS classification criteria (48% accuracy) and 119 formulations using the USEPA classification criteria (43% accuracy). With acute inhalation, the GHS additivity formula misclassified 50 formulations using the GHS classification criteria (61% accuracy) and 34 formulations using the USEPA classification criteria (73% accuracy). For acute dermal toxicity, the GHS additivity formula misclassified 16 formulations using the GHS classification criteria (48% accuracy) and 20 formulations using the USEPA classification criteria (36% accuracy). This data indicates the acute systemic toxicity of many formulations is not the sum of the ingredients' toxicity (additivity); but rather, ingredients in a formulation can interact to result in lower or higher toxicity than predicted by the GHS additivity formula. Copyright © 2018 Elsevier Inc. All rights reserved.

Hybrid Optimization of Object-Based Classification in High-Resolution Images Using Continous ANT Colony Algorithm with Emphasis on Building Detection

NASA Astrophysics Data System (ADS)

Tamimi, E.; Ebadi, H.; Kiani, A.

2017-09-01

Automatic building detection from High Spatial Resolution (HSR) images is one of the most important issues in Remote Sensing (RS). Due to the limited number of spectral bands in HSR images, using other features will lead to improve accuracy. By adding these features, the presence probability of dependent features will be increased, which leads to accuracy reduction. In addition, some parameters should be determined in Support Vector Machine (SVM) classification. Therefore, it is necessary to simultaneously determine classification parameters and select independent features according to image type. Optimization algorithm is an efficient method to solve this problem. On the other hand, pixel-based classification faces several challenges such as producing salt-paper results and high computational time in high dimensional data. Hence, in this paper, a novel method is proposed to optimize object-based SVM classification by applying continuous Ant Colony Optimization (ACO) algorithm. The advantages of the proposed method are relatively high automation level, independency of image scene and type, post processing reduction for building edge reconstruction and accuracy improvement. The proposed method was evaluated by pixel-based SVM and Random Forest (RF) classification in terms of accuracy. In comparison with optimized pixel-based SVM classification, the results showed that the proposed method improved quality factor and overall accuracy by 17% and 10%, respectively. Also, in the proposed method, Kappa coefficient was improved by 6% rather than RF classification. Time processing of the proposed method was relatively low because of unit of image analysis (image object). These showed the superiority of the proposed method in terms of time and accuracy.

Gene masking - a technique to improve accuracy for cancer classification with high dimensionality in microarray data.

PubMed

Saini, Harsh; Lal, Sunil Pranit; Naidu, Vimal Vikash; Pickering, Vincel Wince; Singh, Gurmeet; Tsunoda, Tatsuhiko; Sharma, Alok

2016-12-05

High dimensional feature space generally degrades classification in several applications. In this paper, we propose a strategy called gene masking, in which non-contributing dimensions are heuristically removed from the data to improve classification accuracy. Gene masking is implemented via a binary encoded genetic algorithm that can be integrated seamlessly with classifiers during the training phase of classification to perform feature selection. It can also be used to discriminate between features that contribute most to the classification, thereby, allowing researchers to isolate features that may have special significance. This technique was applied on publicly available datasets whereby it substantially reduced the number of features used for classification while maintaining high accuracies. The proposed technique can be extremely useful in feature selection as it heuristically removes non-contributing features to improve the performance of classifiers.

Comparison of discriminant analysis methods: Application to occupational exposure to particulate matter

NASA Astrophysics Data System (ADS)

Ramos, M. Rosário; Carolino, E.; Viegas, Carla; Viegas, Sandra

2016-06-01

Health effects associated with occupational exposure to particulate matter have been studied by several authors. In this study were selected six industries of five different areas: Cork company 1, Cork company 2, poultry, slaughterhouse for cattle, riding arena and production of animal feed. The measurements tool was a portable device for direct reading. This tool provides information on the particle number concentration for six different diameters, namely 0.3 µm, 0.5 µm, 1 µm, 2.5 µm, 5 µm and 10 µm. The focus on these features is because they might be more closely related with adverse health effects. The aim is to identify the particles that better discriminate the industries, with the ultimate goal of classifying industries regarding potential negative effects on workers' health. Several methods of discriminant analysis were applied to data of occupational exposure to particulate matter and compared with respect to classification accuracy. The selected methods were linear discriminant analyses (LDA); linear quadratic discriminant analysis (QDA), robust linear discriminant analysis with selected estimators (MLE (Maximum Likelihood Estimators), MVE (Minimum Volume Elipsoid), "t", MCD (Minimum Covariance Determinant), MCD-A, MCD-B), multinomial logistic regression and artificial neural networks (ANN). The predictive accuracy of the methods was accessed through a simulation study. ANN yielded the highest rate of classification accuracy in the data set under study. Results indicate that the particle number concentration of diameter size 0.5 µm is the parameter that better discriminates industries.

Analyzing thematic maps and mapping for accuracy

USGS Publications Warehouse

Rosenfield, G.H.

1982-01-01

Two problems which exist while attempting to test the accuracy of thematic maps and mapping are: (1) evaluating the accuracy of thematic content, and (2) evaluating the effects of the variables on thematic mapping. Statistical analysis techniques are applicable to both these problems and include techniques for sampling the data and determining their accuracy. In addition, techniques for hypothesis testing, or inferential statistics, are used when comparing the effects of variables. A comprehensive and valid accuracy test of a classification project, such as thematic mapping from remotely sensed data, includes the following components of statistical analysis: (1) sample design, including the sample distribution, sample size, size of the sample unit, and sampling procedure; and (2) accuracy estimation, including estimation of the variance and confidence limits. Careful consideration must be given to the minimum sample size necessary to validate the accuracy of a given. classification category. The results of an accuracy test are presented in a contingency table sometimes called a classification error matrix. Usually the rows represent the interpretation, and the columns represent the verification. The diagonal elements represent the correct classifications. The remaining elements of the rows represent errors by commission, and the remaining elements of the columns represent the errors of omission. For tests of hypothesis that compare variables, the general practice has been to use only the diagonal elements from several related classification error matrices. These data are arranged in the form of another contingency table. The columns of the table represent the different variables being compared, such as different scales of mapping. The rows represent the blocking characteristics, such as the various categories of classification. The values in the cells of the tables might be the counts of correct classification or the binomial proportions of these counts divided by either the row totals or the column totals from the original classification error matrices. In hypothesis testing, when the results of tests of multiple sample cases prove to be significant, some form of statistical test must be used to separate any results that differ significantly from the others. In the past, many analyses of the data in this error matrix were made by comparing the relative magnitudes of the percentage of correct classifications, for either individual categories, the entire map or both. More rigorous analyses have used data transformations and (or) two-way classification analysis of variance. A more sophisticated step of data analysis techniques would be to use the entire classification error matrices using the methods of discrete multivariate analysis or of multiviariate analysis of variance.

A review of supervised object-based land-cover image classification

NASA Astrophysics Data System (ADS)

Ma, Lei; Li, Manchun; Ma, Xiaoxue; Cheng, Liang; Du, Peijun; Liu, Yongxue

2017-08-01

Object-based image classification for land-cover mapping purposes using remote-sensing imagery has attracted significant attention in recent years. Numerous studies conducted over the past decade have investigated a broad array of sensors, feature selection, classifiers, and other factors of interest. However, these research results have not yet been synthesized to provide coherent guidance on the effect of different supervised object-based land-cover classification processes. In this study, we first construct a database with 28 fields using qualitative and quantitative information extracted from 254 experimental cases described in 173 scientific papers. Second, the results of the meta-analysis are reported, including general characteristics of the studies (e.g., the geographic range of relevant institutes, preferred journals) and the relationships between factors of interest (e.g., spatial resolution and study area or optimal segmentation scale, accuracy and number of targeted classes), especially with respect to the classification accuracy of different sensors, segmentation scale, training set size, supervised classifiers, and land-cover types. Third, useful data on supervised object-based image classification are determined from the meta-analysis. For example, we find that supervised object-based classification is currently experiencing rapid advances, while development of the fuzzy technique is limited in the object-based framework. Furthermore, spatial resolution correlates with the optimal segmentation scale and study area, and Random Forest (RF) shows the best performance in object-based classification. The area-based accuracy assessment method can obtain stable classification performance, and indicates a strong correlation between accuracy and training set size, while the accuracy of the point-based method is likely to be unstable due to mixed objects. In addition, the overall accuracy benefits from higher spatial resolution images (e.g., unmanned aerial vehicle) or agricultural sites where it also correlates with the number of targeted classes. More than 95.6% of studies involve an area less than 300 ha, and the spatial resolution of images is predominantly between 0 and 2 m. Furthermore, we identify some methods that may advance supervised object-based image classification. For example, deep learning and type-2 fuzzy techniques may further improve classification accuracy. Lastly, scientists are strongly encouraged to report results of uncertainty studies to further explore the effects of varied factors on supervised object-based image classification.

Localized thin-section CT with radiomics feature extraction and machine learning to classify early-detected pulmonary nodules from lung cancer screening

NASA Astrophysics Data System (ADS)

Tu, Shu-Ju; Wang, Chih-Wei; Pan, Kuang-Tse; Wu, Yi-Cheng; Wu, Chen-Te

2018-03-01

Lung cancer screening aims to detect small pulmonary nodules and decrease the mortality rate of those affected. However, studies from large-scale clinical trials of lung cancer screening have shown that the false-positive rate is high and positive predictive value is low. To address these problems, a technical approach is greatly needed for accurate malignancy differentiation among these early-detected nodules. We studied the clinical feasibility of an additional protocol of localized thin-section CT for further assessment on recalled patients from lung cancer screening tests. Our approach of localized thin-section CT was integrated with radiomics features extraction and machine learning classification which was supervised by pathological diagnosis. Localized thin-section CT images of 122 nodules were retrospectively reviewed and 374 radiomics features were extracted. In this study, 48 nodules were benign and 74 malignant. There were nine patients with multiple nodules and four with synchronous multiple malignant nodules. Different machine learning classifiers with a stratified ten-fold cross-validation were used and repeated 100 times to evaluate classification accuracy. Of the image features extracted from the thin-section CT images, 238 (64%) were useful in differentiating between benign and malignant nodules. These useful features include CT density (p  =  0.002 518), sigma (p  =  0.002 781), uniformity (p  =  0.032 41), and entropy (p  =  0.006 685). The highest classification accuracy was 79% by the logistic classifier. The performance metrics of this logistic classification model was 0.80 for the positive predictive value, 0.36 for the false-positive rate, and 0.80 for the area under the receiver operating characteristic curve. Our approach of direct risk classification supervised by the pathological diagnosis with localized thin-section CT and radiomics feature extraction may support clinical physicians in determining truly malignant nodules and therefore reduce problems in lung cancer screening.

Focal liver lesions segmentation and classification in nonenhanced T2-weighted MRI.

PubMed

Gatos, Ilias; Tsantis, Stavros; Karamesini, Maria; Spiliopoulos, Stavros; Karnabatidis, Dimitris; Hazle, John D; Kagadis, George C

2017-07-01

To automatically segment and classify focal liver lesions (FLLs) on nonenhanced T2-weighted magnetic resonance imaging (MRI) scans using a computer-aided diagnosis (CAD) algorithm. 71 FLLs (30 benign lesions, 19 hepatocellular carcinomas, and 22 metastases) on T2-weighted MRI scans were delineated by the proposed CAD scheme. The FLL segmentation procedure involved wavelet multiscale analysis to extract accurate edge information and mean intensity values for consecutive edges computed using horizontal and vertical analysis that were fed into the subsequent fuzzy C-means algorithm for final FLL border extraction. Texture information for each extracted lesion was derived using 42 first- and second-order textural features from grayscale value histogram, co-occurrence, and run-length matrices. Twelve morphological features were also extracted to capture any shape differentiation between classes. Feature selection was performed with stepwise multilinear regression analysis that led to a reduced feature subset. A multiclass Probabilistic Neural Network (PNN) classifier was then designed and used for lesion classification. PNN model evaluation was performed using the leave-one-out (LOO) method and receiver operating characteristic (ROC) curve analysis. The mean overlap between the automatically segmented FLLs and the manual segmentations performed by radiologists was 0.91 ± 0.12. The highest classification accuracies in the PNN model for the benign, hepatocellular carcinoma, and metastatic FLLs were 94.1%, 91.4%, and 94.1%, respectively, with sensitivity/specificity values of 90%/97.3%, 89.5%/92.2%, and 90.9%/95.6% respectively. The overall classification accuracy for the proposed system was 90.1%. Our diagnostic system using sophisticated FLL segmentation and classification algorithms is a powerful tool for routine clinical MRI-based liver evaluation and can be a supplement to contrast-enhanced MRI to prevent unnecessary invasive procedures. © 2017 American Association of Physicists in Medicine.

Localized thin-section CT with radiomics feature extraction and machine learning to classify early-detected pulmonary nodules from lung cancer screening.

PubMed

Tu, Shu-Ju; Wang, Chih-Wei; Pan, Kuang-Tse; Wu, Yi-Cheng; Wu, Chen-Te

2018-03-14

Lung cancer screening aims to detect small pulmonary nodules and decrease the mortality rate of those affected. However, studies from large-scale clinical trials of lung cancer screening have shown that the false-positive rate is high and positive predictive value is low. To address these problems, a technical approach is greatly needed for accurate malignancy differentiation among these early-detected nodules. We studied the clinical feasibility of an additional protocol of localized thin-section CT for further assessment on recalled patients from lung cancer screening tests. Our approach of localized thin-section CT was integrated with radiomics features extraction and machine learning classification which was supervised by pathological diagnosis. Localized thin-section CT images of 122 nodules were retrospectively reviewed and 374 radiomics features were extracted. In this study, 48 nodules were benign and 74 malignant. There were nine patients with multiple nodules and four with synchronous multiple malignant nodules. Different machine learning classifiers with a stratified ten-fold cross-validation were used and repeated 100 times to evaluate classification accuracy. Of the image features extracted from the thin-section CT images, 238 (64%) were useful in differentiating between benign and malignant nodules. These useful features include CT density (p  =  0.002 518), sigma (p  =  0.002 781), uniformity (p  =  0.032 41), and entropy (p  =  0.006 685). The highest classification accuracy was 79% by the logistic classifier. The performance metrics of this logistic classification model was 0.80 for the positive predictive value, 0.36 for the false-positive rate, and 0.80 for the area under the receiver operating characteristic curve. Our approach of direct risk classification supervised by the pathological diagnosis with localized thin-section CT and radiomics feature extraction may support clinical physicians in determining truly malignant nodules and therefore reduce problems in lung cancer screening.

Reliability, Validity, and Classification Accuracy of the DSM-5 Diagnostic Criteria for Gambling Disorder and Comparison to DSM-IV.

PubMed

Stinchfield, Randy; McCready, John; Turner, Nigel E; Jimenez-Murcia, Susana; Petry, Nancy M; Grant, Jon; Welte, John; Chapman, Heather; Winters, Ken C

2016-09-01

The DSM-5 was published in 2013 and it included two substantive revisions for gambling disorder (GD). These changes are the reduction in the threshold from five to four criteria and elimination of the illegal activities criterion. The purpose of this study was to twofold. First, to assess the reliability, validity and classification accuracy of the DSM-5 diagnostic criteria for GD. Second, to compare the DSM-5-DSM-IV on reliability, validity, and classification accuracy, including an examination of the effect of the elimination of the illegal acts criterion on diagnostic accuracy. To compare DSM-5 and DSM-IV, eight datasets from three different countries (Canada, USA, and Spain; total N = 3247) were used. All datasets were based on similar research methods. Participants were recruited from outpatient gambling treatment services to represent the group with a GD and from the community to represent the group without a GD. All participants were administered a standardized measure of diagnostic criteria. The DSM-5 yielded satisfactory reliability, validity and classification accuracy. In comparing the DSM-5 to the DSM-IV, most comparisons of reliability, validity and classification accuracy showed more similarities than differences. There was evidence of modest improvements in classification accuracy for DSM-5 over DSM-IV, particularly in reduction of false negative errors. This reduction in false negative errors was largely a function of lowering the cut score from five to four and this revision is an improvement over DSM-IV. From a statistical standpoint, eliminating the illegal acts criterion did not make a significant impact on diagnostic accuracy. From a clinical standpoint, illegal acts can still be addressed in the context of the DSM-5 criterion of lying to others.

Practical Issues in Estimating Classification Accuracy and Consistency with R Package cacIRT

ERIC Educational Resources Information Center

Lathrop, Quinn N.

2015-01-01

There are two main lines of research in estimating classification accuracy (CA) and classification consistency (CC) under Item Response Theory (IRT). The R package cacIRT provides computer implementations of both approaches in an accessible and unified framework. Even with available implementations, there remains decisions a researcher faces when…

Variance estimates and confidence intervals for the Kappa measure of classification accuracy

Treesearch

M. A. Kalkhan; R. M. Reich; R. L. Czaplewski

1997-01-01

The Kappa statistic is frequently used to characterize the results of an accuracy assessment used to evaluate land use and land cover classifications obtained by remotely sensed data. This statistic allows comparisons of alternative sampling designs, classification algorithms, photo-interpreters, and so forth. In order to make these comparisons, it is...

HEp-2 cell image classification method based on very deep convolutional networks with small datasets

NASA Astrophysics Data System (ADS)

Lu, Mengchi; Gao, Long; Guo, Xifeng; Liu, Qiang; Yin, Jianping

2017-07-01

Human Epithelial-2 (HEp-2) cell images staining patterns classification have been widely used to identify autoimmune diseases by the anti-Nuclear antibodies (ANA) test in the Indirect Immunofluorescence (IIF) protocol. Because manual test is time consuming, subjective and labor intensive, image-based Computer Aided Diagnosis (CAD) systems for HEp-2 cell classification are developing. However, methods proposed recently are mostly manual features extraction with low accuracy. Besides, the scale of available benchmark datasets is small, which does not exactly suitable for using deep learning methods. This issue will influence the accuracy of cell classification directly even after data augmentation. To address these issues, this paper presents a high accuracy automatic HEp-2 cell classification method with small datasets, by utilizing very deep convolutional networks (VGGNet). Specifically, the proposed method consists of three main phases, namely image preprocessing, feature extraction and classification. Moreover, an improved VGGNet is presented to address the challenges of small-scale datasets. Experimental results over two benchmark datasets demonstrate that the proposed method achieves superior performance in terms of accuracy compared with existing methods.

On-line analysis of algae in water by discrete three-dimensional fluorescence spectroscopy.

PubMed

Zhao, Nanjing; Zhang, Xiaoling; Yin, Gaofang; Yang, Ruifang; Hu, Li; Chen, Shuang; Liu, Jianguo; Liu, Wenqing

2018-03-19

In view of the problem of the on-line measurement of algae classification, a method of algae classification and concentration determination based on the discrete three-dimensional fluorescence spectra was studied in this work. The discrete three-dimensional fluorescence spectra of twelve common species of algae belonging to five categories were analyzed, the discrete three-dimensional standard spectra of five categories were built, and the recognition, classification and concentration prediction of algae categories were realized by the discrete three-dimensional fluorescence spectra coupled with non-negative weighted least squares linear regression analysis. The results show that similarities between discrete three-dimensional standard spectra of different categories were reduced and the accuracies of recognition, classification and concentration prediction of the algae categories were significantly improved. By comparing with that of the chlorophyll a fluorescence excitation spectra method, the recognition accuracy rate in pure samples by discrete three-dimensional fluorescence spectra is improved 1.38%, and the recovery rate and classification accuracy in pure diatom samples 34.1% and 46.8%, respectively; the recognition accuracy rate of mixed samples by discrete-three dimensional fluorescence spectra is enhanced by 26.1%, the recovery rate of mixed samples with Chlorophyta 37.8%, and the classification accuracy of mixed samples with diatoms 54.6%.

Comparing supervised and unsupervised multiresolution segmentation approaches for extracting buildings from very high resolution imagery.

PubMed

Belgiu, Mariana; Dr Guţ, Lucian

2014-10-01

Although multiresolution segmentation (MRS) is a powerful technique for dealing with very high resolution imagery, some of the image objects that it generates do not match the geometries of the target objects, which reduces the classification accuracy. MRS can, however, be guided to produce results that approach the desired object geometry using either supervised or unsupervised approaches. Although some studies have suggested that a supervised approach is preferable, there has been no comparative evaluation of these two approaches. Therefore, in this study, we have compared supervised and unsupervised approaches to MRS. One supervised and two unsupervised segmentation methods were tested on three areas using QuickBird and WorldView-2 satellite imagery. The results were assessed using both segmentation evaluation methods and an accuracy assessment of the resulting building classifications. Thus, differences in the geometries of the image objects and in the potential to achieve satisfactory thematic accuracies were evaluated. The two approaches yielded remarkably similar classification results, with overall accuracies ranging from 82% to 86%. The performance of one of the unsupervised methods was unexpectedly similar to that of the supervised method; they identified almost identical scale parameters as being optimal for segmenting buildings, resulting in very similar geometries for the resulting image objects. The second unsupervised method produced very different image objects from the supervised method, but their classification accuracies were still very similar. The latter result was unexpected because, contrary to previously published findings, it suggests a high degree of independence between the segmentation results and classification accuracy. The results of this study have two important implications. The first is that object-based image analysis can be automated without sacrificing classification accuracy, and the second is that the previously accepted idea that classification is dependent on segmentation is challenged by our unexpected results, casting doubt on the value of pursuing 'optimal segmentation'. Our results rather suggest that as long as under-segmentation remains at acceptable levels, imperfections in segmentation can be ruled out, so that a high level of classification accuracy can still be achieved.

Exploring the impact of wavelet-based denoising in the classification of remote sensing hyperspectral images

NASA Astrophysics Data System (ADS)

Quesada-Barriuso, Pablo; Heras, Dora B.; Argüello, Francisco

2016-10-01

The classification of remote sensing hyperspectral images for land cover applications is a very intensive topic. In the case of supervised classification, Support Vector Machines (SVMs) play a dominant role. Recently, the Extreme Learning Machine algorithm (ELM) has been extensively used. The classification scheme previously published by the authors, and called WT-EMP, introduces spatial information in the classification process by means of an Extended Morphological Profile (EMP) that is created from features extracted by wavelets. In addition, the hyperspectral image is denoised in the 2-D spatial domain, also using wavelets and it is joined to the EMP via a stacked vector. In this paper, the scheme is improved achieving two goals. The first one is to reduce the classification time while preserving the accuracy of the classification by using ELM instead of SVM. The second one is to improve the accuracy results by performing not only a 2-D denoising for every spectral band, but also a previous additional 1-D spectral signature denoising applied to each pixel vector of the image. For each denoising the image is transformed by applying a 1-D or 2-D wavelet transform, and then a NeighShrink thresholding is applied. Improvements in terms of classification accuracy are obtained, especially for images with close regions in the classification reference map, because in these cases the accuracy of the classification in the edges between classes is more relevant.

Comparing Features for Classification of MEG Responses to Motor Imagery

PubMed Central

Halme, Hanna-Leena; Parkkonen, Lauri

2016-01-01

Background Motor imagery (MI) with real-time neurofeedback could be a viable approach, e.g., in rehabilitation of cerebral stroke. Magnetoencephalography (MEG) noninvasively measures electric brain activity at high temporal resolution and is well-suited for recording oscillatory brain signals. MI is known to modulate 10- and 20-Hz oscillations in the somatomotor system. In order to provide accurate feedback to the subject, the most relevant MI-related features should be extracted from MEG data. In this study, we evaluated several MEG signal features for discriminating between left- and right-hand MI and between MI and rest. Methods MEG was measured from nine healthy participants imagining either left- or right-hand finger tapping according to visual cues. Data preprocessing, feature extraction and classification were performed offline. The evaluated MI-related features were power spectral density (PSD), Morlet wavelets, short-time Fourier transform (STFT), common spatial patterns (CSP), filter-bank common spatial patterns (FBCSP), spatio—spectral decomposition (SSD), and combined SSD+CSP, CSP+PSD, CSP+Morlet, and CSP+STFT. We also compared four classifiers applied to single trials using 5-fold cross-validation for evaluating the classification accuracy and its possible dependence on the classification algorithm. In addition, we estimated the inter-session left-vs-right accuracy for each subject. Results The SSD+CSP combination yielded the best accuracy in both left-vs-right (mean 73.7%) and MI-vs-rest (mean 81.3%) classification. CSP+Morlet yielded the best mean accuracy in inter-session left-vs-right classification (mean 69.1%). There were large inter-subject differences in classification accuracy, and the level of the 20-Hz suppression correlated significantly with the subjective MI-vs-rest accuracy. Selection of the classification algorithm had only a minor effect on the results. Conclusions We obtained good accuracy in sensor-level decoding of MI from single-trial MEG data. Feature extraction methods utilizing both the spatial and spectral profile of MI-related signals provided the best classification results, suggesting good performance of these methods in an online MEG neurofeedback system. PMID:27992574

Differentiation of plant age in grasses using remote sensing

NASA Astrophysics Data System (ADS)

Knox, Nichola M.; Skidmore, Andrew K.; van der Werff, Harald M. A.; Groen, Thomas A.; de Boer, Willem F.; Prins, Herbert H. T.; Kohi, Edward; Peel, Mike

2013-10-01

Phenological or plant age classification across a landscape allows for examination of micro-topographical effects on plant growth, improvement in the accuracy of species discrimination, and will improve our understanding of the spatial variation in plant growth. In this paper six vegetation indices used in phenological studies (including the newly proposed PhIX index) were analysed for their ability to statistically differentiate grasses of different ages in the sequence of their development. Spectra of grasses of different ages were collected from a greenhouse study. These were used to determine if NDVI, NDWI, CAI, EVI, EVI2 and the newly proposed PhIX index could sequentially discriminate grasses of different ages, and subsequently classify grasses into their respective age category. The PhIX index was defined as: (AVNIRn+log(ASWIR2n))/(AVNIRn-log(ASWIR2n)), where AVNIRn and ASWIR2n are the respective normalised areas under the continuum removed reflectance curve within the VNIR (500-800 nm) and SWIR2 (2000-2210 nm) regions. The PhIX index was found to produce the highest phenological classification accuracy (Overall Accuracy: 79%, and Kappa Accuracy: 75%) and similar to the NDVI, EVI and EVI2 indices it statistically sequentially separates out the developmental age classes. Discrimination between seedling and dormant age classes and the adult and flowering classes was problematic for most of the tested indices. Combining information from the visible near infrared (VNIR) and shortwave infrared region (SWIR) region into a single phenological index captures the phenological changes associated with plant pigments and the ligno-cellulose absorption feature, providing a robust method to discriminate the age classes of grasses. This work provides a valuable contribution into mapping spatial variation and monitoring plant growth across savanna and grassland ecosystems.

Spatiotemporal Beamforming: A Transparent and Unified Decoding Approach to Synchronous Visual Brain-Computer Interfacing.

PubMed

Wittevrongel, Benjamin; Van Hulle, Marc M

2017-01-01

Brain-Computer Interfaces (BCIs) decode brain activity with the aim to establish a direct communication channel with an external device. Albeit they have been hailed to (re-)establish communication in persons suffering from severe motor- and/or communication disabilities, only recently BCI applications have been challenging other assistive technologies. Owing to their considerably increased performance and the advent of affordable technological solutions, BCI technology is expected to trigger a paradigm shift not only in assistive technology but also in the way we will interface with technology. However, the flipside of the quest for accuracy and speed is most evident in EEG-based visual BCI where it has led to a gamut of increasingly complex classifiers, tailored to the needs of specific stimulation paradigms and use contexts. In this contribution, we argue that spatiotemporal beamforming can serve several synchronous visual BCI paradigms. We demonstrate this for three popular visual paradigms even without attempting to optimizing their electrode sets. For each selectable target, a spatiotemporal beamformer is applied to assess whether the corresponding signal-of-interest is present in the preprocessed multichannel EEG signals. The target with the highest beamformer output is then selected by the decoder (maximum selection). In addition to this simple selection rule, we also investigated whether interactions between beamformer outputs could be employed to increase accuracy by combining the outputs for all targets into a feature vector and applying three common classification algorithms. The results show that the accuracy of spatiotemporal beamforming with maximum selection is at par with that of the classification algorithms and interactions between beamformer outputs do not further improve that accuracy.

Automatic detection of osteoporosis based on hybrid genetic swarm fuzzy classifier approaches

PubMed Central

Kavitha, Muthu Subash; Ganesh Kumar, Pugalendhi; Park, Soon-Yong; Huh, Kyung-Hoe; Heo, Min-Suk; Kurita, Takio; Asano, Akira; An, Seo-Yong

2016-01-01

Objectives: This study proposed a new automated screening system based on a hybrid genetic swarm fuzzy (GSF) classifier using digital dental panoramic radiographs to diagnose females with a low bone mineral density (BMD) or osteoporosis. Methods: The geometrical attributes of both the mandibular cortical bone and trabecular bone were acquired using previously developed software. Designing an automated system for osteoporosis screening involved partitioning of the input attributes to generate an initial membership function (MF) and a rule set (RS), classification using a fuzzy inference system and optimization of the generated MF and RS using the genetic swarm algorithm. Fivefold cross-validation (5-FCV) was used to estimate the classification accuracy of the hybrid GSF classifier. The performance of the hybrid GSF classifier has been further compared with that of individual genetic algorithm and particle swarm optimization fuzzy classifiers. Results: Proposed hybrid GSF classifier in identifying low BMD or osteoporosis at the lumbar spine and femoral neck BMD was evaluated. The sensitivity, specificity and accuracy of the hybrid GSF with optimized MF and RS in identifying females with a low BMD were 95.3%, 94.7% and 96.01%, respectively, at the lumbar spine and 99.1%, 98.4% and 98.9%, respectively, at the femoral neck BMD. The diagnostic performance of the proposed system with femoral neck BMD was 0.986 with a confidence interval of 0.942–0.998. The highest mean accuracy using 5-FCV was 97.9% with femoral neck BMD. Conclusions: The combination of high accuracy along with its interpretation ability makes this proposed automatic system using hybrid GSF classifier capable of identifying a large proportion of undetected low BMD or osteoporosis at its early stage. PMID:27186991

A simple and robust classification tree for differentiation between benign and malignant lesions in MR-mammography.

PubMed

Baltzer, Pascal A T; Dietzel, Matthias; Kaiser, Werner A

2013-08-01

In the face of multiple available diagnostic criteria in MR-mammography (MRM), a practical algorithm for lesion classification is needed. Such an algorithm should be as simple as possible and include only important independent lesion features to differentiate benign from malignant lesions. This investigation aimed to develop a simple classification tree for differential diagnosis in MRM. A total of 1,084 lesions in standardised MRM with subsequent histological verification (648 malignant, 436 benign) were investigated. Seventeen lesion criteria were assessed by 2 readers in consensus. Classification analysis was performed using the chi-squared automatic interaction detection (CHAID) method. Results include the probability for malignancy for every descriptor combination in the classification tree. A classification tree incorporating 5 lesion descriptors with a depth of 3 ramifications (1, root sign; 2, delayed enhancement pattern; 3, border, internal enhancement and oedema) was calculated. Of all 1,084 lesions, 262 (40.4 %) and 106 (24.3 %) could be classified as malignant and benign with an accuracy above 95 %, respectively. Overall diagnostic accuracy was 88.4 %. The classification algorithm reduced the number of categorical descriptors from 17 to 5 (29.4 %), resulting in a high classification accuracy. More than one third of all lesions could be classified with accuracy above 95 %. • A practical algorithm has been developed to classify lesions found in MR-mammography. • A simple decision tree consisting of five criteria reaches high accuracy of 88.4 %. • Unique to this approach, each classification is associated with a diagnostic certainty. • Diagnostic certainty of greater than 95 % is achieved in 34 % of all cases.

Improving EEG-Based Driver Fatigue Classification Using Sparse-Deep Belief Networks.

PubMed

Chai, Rifai; Ling, Sai Ho; San, Phyo Phyo; Naik, Ganesh R; Nguyen, Tuan N; Tran, Yvonne; Craig, Ashley; Nguyen, Hung T

2017-01-01

This paper presents an improvement of classification performance for electroencephalography (EEG)-based driver fatigue classification between fatigue and alert states with the data collected from 43 participants. The system employs autoregressive (AR) modeling as the features extraction algorithm, and sparse-deep belief networks (sparse-DBN) as the classification algorithm. Compared to other classifiers, sparse-DBN is a semi supervised learning method which combines unsupervised learning for modeling features in the pre-training layer and supervised learning for classification in the following layer. The sparsity in sparse-DBN is achieved with a regularization term that penalizes a deviation of the expected activation of hidden units from a fixed low-level prevents the network from overfitting and is able to learn low-level structures as well as high-level structures. For comparison, the artificial neural networks (ANN), Bayesian neural networks (BNN), and original deep belief networks (DBN) classifiers are used. The classification results show that using AR feature extractor and DBN classifiers, the classification performance achieves an improved classification performance with a of sensitivity of 90.8%, a specificity of 90.4%, an accuracy of 90.6%, and an area under the receiver operating curve (AUROC) of 0.94 compared to ANN (sensitivity at 80.8%, specificity at 77.8%, accuracy at 79.3% with AUC-ROC of 0.83) and BNN classifiers (sensitivity at 84.3%, specificity at 83%, accuracy at 83.6% with AUROC of 0.87). Using the sparse-DBN classifier, the classification performance improved further with sensitivity of 93.9%, a specificity of 92.3%, and an accuracy of 93.1% with AUROC of 0.96. Overall, the sparse-DBN classifier improved accuracy by 13.8, 9.5, and 2.5% over ANN, BNN, and DBN classifiers, respectively.

Improving EEG-Based Driver Fatigue Classification Using Sparse-Deep Belief Networks

PubMed Central

Chai, Rifai; Ling, Sai Ho; San, Phyo Phyo; Naik, Ganesh R.; Nguyen, Tuan N.; Tran, Yvonne; Craig, Ashley; Nguyen, Hung T.

2017-01-01

This paper presents an improvement of classification performance for electroencephalography (EEG)-based driver fatigue classification between fatigue and alert states with the data collected from 43 participants. The system employs autoregressive (AR) modeling as the features extraction algorithm, and sparse-deep belief networks (sparse-DBN) as the classification algorithm. Compared to other classifiers, sparse-DBN is a semi supervised learning method which combines unsupervised learning for modeling features in the pre-training layer and supervised learning for classification in the following layer. The sparsity in sparse-DBN is achieved with a regularization term that penalizes a deviation of the expected activation of hidden units from a fixed low-level prevents the network from overfitting and is able to learn low-level structures as well as high-level structures. For comparison, the artificial neural networks (ANN), Bayesian neural networks (BNN), and original deep belief networks (DBN) classifiers are used. The classification results show that using AR feature extractor and DBN classifiers, the classification performance achieves an improved classification performance with a of sensitivity of 90.8%, a specificity of 90.4%, an accuracy of 90.6%, and an area under the receiver operating curve (AUROC) of 0.94 compared to ANN (sensitivity at 80.8%, specificity at 77.8%, accuracy at 79.3% with AUC-ROC of 0.83) and BNN classifiers (sensitivity at 84.3%, specificity at 83%, accuracy at 83.6% with AUROC of 0.87). Using the sparse-DBN classifier, the classification performance improved further with sensitivity of 93.9%, a specificity of 92.3%, and an accuracy of 93.1% with AUROC of 0.96. Overall, the sparse-DBN classifier improved accuracy by 13.8, 9.5, and 2.5% over ANN, BNN, and DBN classifiers, respectively. PMID:28326009

[Combining speech sample and feature bilateral selection algorithm for classification of Parkinson's disease].

PubMed

Zhang, Xiaoheng; Wang, Lirui; Cao, Yao; Wang, Pin; Zhang, Cheng; Yang, Liuyang; Li, Yongming; Zhang, Yanling; Cheng, Oumei

2018-02-01

Diagnosis of Parkinson's disease (PD) based on speech data has been proved to be an effective way in recent years. However, current researches just care about the feature extraction and classifier design, and do not consider the instance selection. Former research by authors showed that the instance selection can lead to improvement on classification accuracy. However, no attention is paid on the relationship between speech sample and feature until now. Therefore, a new diagnosis algorithm of PD is proposed in this paper by simultaneously selecting speech sample and feature based on relevant feature weighting algorithm and multiple kernel method, so as to find their synergy effects, thereby improving classification accuracy. Experimental results showed that this proposed algorithm obtained apparent improvement on classification accuracy. It can obtain mean classification accuracy of 82.5%, which was 30.5% higher than the relevant algorithm. Besides, the proposed algorithm detected the synergy effects of speech sample and feature, which is valuable for speech marker extraction.

A neural network approach to cloud classification

NASA Technical Reports Server (NTRS)

Lee, Jonathan; Weger, Ronald C.; Sengupta, Sailes K.; Welch, Ronald M.

1990-01-01

It is shown that, using high-spatial-resolution data, very high cloud classification accuracies can be obtained with a neural network approach. A texture-based neural network classifier using only single-channel visible Landsat MSS imagery achieves an overall cloud identification accuracy of 93 percent. Cirrus can be distinguished from boundary layer cloudiness with an accuracy of 96 percent, without the use of an infrared channel. Stratocumulus is retrieved with an accuracy of 92 percent, cumulus at 90 percent. The use of the neural network does not improve cirrus classification accuracy. Rather, its main effect is in the improved separation between stratocumulus and cumulus cloudiness. While most cloud classification algorithms rely on linear parametric schemes, the present study is based on a nonlinear, nonparametric four-layer neural network approach. A three-layer neural network architecture, the nonparametric K-nearest neighbor approach, and the linear stepwise discriminant analysis procedure are compared. A significant finding is that significantly higher accuracies are attained with the nonparametric approaches using only 20 percent of the database as training data, compared to 67 percent of the database in the linear approach.

IMPROVING THE ACCURACY OF HISTORIC SATELLITE IMAGE CLASSIFICATION BY COMBINING LOW-RESOLUTION MULTISPECTRAL DATA WITH HIGH-RESOLUTION PANCHROMATIC DATA

DOE Office of Scientific and Technical Information (OSTI.GOV)

Getman, Daniel J

2008-01-01

Many attempts to observe changes in terrestrial systems over time would be significantly enhanced if it were possible to improve the accuracy of classifications of low-resolution historic satellite data. In an effort to examine improving the accuracy of historic satellite image classification by combining satellite and air photo data, two experiments were undertaken in which low-resolution multispectral data and high-resolution panchromatic data were combined and then classified using the ECHO spectral-spatial image classification algorithm and the Maximum Likelihood technique. The multispectral data consisted of 6 multispectral channels (30-meter pixel resolution) from Landsat 7. These data were augmented with panchromatic datamore » (15m pixel resolution) from Landsat 7 in the first experiment, and with a mosaic of digital aerial photography (1m pixel resolution) in the second. The addition of the Landsat 7 panchromatic data provided a significant improvement in the accuracy of classifications made using the ECHO algorithm. Although the inclusion of aerial photography provided an improvement in accuracy, this improvement was only statistically significant at a 40-60% level. These results suggest that once error levels associated with combining aerial photography and multispectral satellite data are reduced, this approach has the potential to significantly enhance the precision and accuracy of classifications made using historic remotely sensed data, as a way to extend the time range of efforts to track temporal changes in terrestrial systems.« less

Data mining methods in the prediction of Dementia: A real-data comparison of the accuracy, sensitivity and specificity of linear discriminant analysis, logistic regression, neural networks, support vector machines, classification trees and random forests

PubMed Central

2011-01-01

Background Dementia and cognitive impairment associated with aging are a major medical and social concern. Neuropsychological testing is a key element in the diagnostic procedures of Mild Cognitive Impairment (MCI), but has presently a limited value in the prediction of progression to dementia. We advance the hypothesis that newer statistical classification methods derived from data mining and machine learning methods like Neural Networks, Support Vector Machines and Random Forests can improve accuracy, sensitivity and specificity of predictions obtained from neuropsychological testing. Seven non parametric classifiers derived from data mining methods (Multilayer Perceptrons Neural Networks, Radial Basis Function Neural Networks, Support Vector Machines, CART, CHAID and QUEST Classification Trees and Random Forests) were compared to three traditional classifiers (Linear Discriminant Analysis, Quadratic Discriminant Analysis and Logistic Regression) in terms of overall classification accuracy, specificity, sensitivity, Area under the ROC curve and Press'Q. Model predictors were 10 neuropsychological tests currently used in the diagnosis of dementia. Statistical distributions of classification parameters obtained from a 5-fold cross-validation were compared using the Friedman's nonparametric test. Results Press' Q test showed that all classifiers performed better than chance alone (p < 0.05). Support Vector Machines showed the larger overall classification accuracy (Median (Me) = 0.76) an area under the ROC (Me = 0.90). However this method showed high specificity (Me = 1.0) but low sensitivity (Me = 0.3). Random Forest ranked second in overall accuracy (Me = 0.73) with high area under the ROC (Me = 0.73) specificity (Me = 0.73) and sensitivity (Me = 0.64). Linear Discriminant Analysis also showed acceptable overall accuracy (Me = 0.66), with acceptable area under the ROC (Me = 0.72) specificity (Me = 0.66) and sensitivity (Me = 0.64). The remaining classifiers showed overall classification accuracy above a median value of 0.63, but for most sensitivity was around or even lower than a median value of 0.5. Conclusions When taking into account sensitivity, specificity and overall classification accuracy Random Forests and Linear Discriminant analysis rank first among all the classifiers tested in prediction of dementia using several neuropsychological tests. These methods may be used to improve accuracy, sensitivity and specificity of Dementia predictions from neuropsychological testing. PMID:21849043

Does Maximizing Information at the Cut Score Always Maximize Classification Accuracy and Consistency?

ERIC Educational Resources Information Center

Wyse, Adam E.; Babcock, Ben

2016-01-01

A common suggestion made in the psychometric literature for fixed-length classification tests is that one should design tests so that they have maximum information at the cut score. Designing tests in this way is believed to maximize the classification accuracy and consistency of the assessment. This article uses simulated examples to illustrate…

Improved Classification of Lung Cancer Using Radial Basis Function Neural Network with Affine Transforms of Voss Representation.

PubMed

Adetiba, Emmanuel; Olugbara, Oludayo O

2015-01-01

Lung cancer is one of the diseases responsible for a large number of cancer related death cases worldwide. The recommended standard for screening and early detection of lung cancer is the low dose computed tomography. However, many patients diagnosed die within one year, which makes it essential to find alternative approaches for screening and early detection of lung cancer. We present computational methods that can be implemented in a functional multi-genomic system for classification, screening and early detection of lung cancer victims. Samples of top ten biomarker genes previously reported to have the highest frequency of lung cancer mutations and sequences of normal biomarker genes were respectively collected from the COSMIC and NCBI databases to validate the computational methods. Experiments were performed based on the combinations of Z-curve and tetrahedron affine transforms, Histogram of Oriented Gradient (HOG), Multilayer perceptron and Gaussian Radial Basis Function (RBF) neural networks to obtain an appropriate combination of computational methods to achieve improved classification of lung cancer biomarker genes. Results show that a combination of affine transforms of Voss representation, HOG genomic features and Gaussian RBF neural network perceptibly improves classification accuracy, specificity and sensitivity of lung cancer biomarker genes as well as achieving low mean square error.

Visual modifications on the P300 speller BCI paradigm

NASA Astrophysics Data System (ADS)

Salvaris, M.; Sepulveda, F.

2009-08-01

The best known P300 speller brain-computer interface (BCI) paradigm is the Farwell and Donchin paradigm. In this paper, various changes to the visual aspects of this protocol are explored as well as their effects on classification. Changes to the dimensions of the symbols, the distance between the symbols and the colours used were tested. The purpose of the present work was not to achieve the highest possible accuracy results, but to ascertain whether these simple modifications to the visual protocol will provide classification differences between them and what these differences will be. Eight subjects were used, with each subject carrying out a total of six different experiments. In each experiment, the user spelt a total of 39 characters. Two types of classifiers were trained and tested to determine whether the results were classifier dependant. These were a support vector machine (SVM) with a radial basis function (RBF) kernel and Fisher's linear discriminant (FLD). The single-trial classification results and multiple-trial classification results were recorded and compared. Although no visual protocol was the best for all subjects, the best performances, across both classifiers, were obtained with the white background (WB) visual protocol. The worst performance was obtained with the small symbol size (SSS) visual protocol.

Global land cover mapping: a review and uncertainty analysis

USGS Publications Warehouse

Congalton, Russell G.; Gu, Jianyu; Yadav, Kamini; Thenkabail, Prasad S.; Ozdogan, Mutlu

2014-01-01

Given the advances in remotely sensed imagery and associated technologies, several global land cover maps have been produced in recent times including IGBP DISCover, UMD Land Cover, Global Land Cover 2000 and GlobCover 2009. However, the utility of these maps for specific applications has often been hampered due to considerable amounts of uncertainties and inconsistencies. A thorough review of these global land cover projects including evaluating the sources of error and uncertainty is prudent and enlightening. Therefore, this paper describes our work in which we compared, summarized and conducted an uncertainty analysis of the four global land cover mapping projects using an error budget approach. The results showed that the classification scheme and the validation methodology had the highest error contribution and implementation priority. A comparison of the classification schemes showed that there are many inconsistencies between the definitions of the map classes. This is especially true for the mixed type classes for which thresholds vary for the attributes/discriminators used in the classification process. Examination of these four global mapping projects provided quite a few important lessons for the future global mapping projects including the need for clear and uniform definitions of the classification scheme and an efficient, practical, and valid design of the accuracy assessment.

Classification of fresh and frozen-thawed pork muscles using visible and near infrared hyperspectral imaging and textural analysis.

PubMed

Pu, Hongbin; Sun, Da-Wen; Ma, Ji; Cheng, Jun-Hu

2015-01-01

The potential of visible and near infrared hyperspectral imaging was investigated as a rapid and nondestructive technique for classifying fresh and frozen-thawed meats by integrating critical spectral and image features extracted from hyperspectral images in the region of 400-1000 nm. Six feature wavelengths (400, 446, 477, 516, 592 and 686 nm) were identified using uninformative variable elimination and successive projections algorithm. Image textural features of the principal component images from hyperspectral images were obtained using histogram statistics (HS), gray level co-occurrence matrix (GLCM) and gray level-gradient co-occurrence matrix (GLGCM). By these spectral and textural features, probabilistic neural network (PNN) models for classification of fresh and frozen-thawed pork meats were established. Compared with the models using the optimum wavelengths only, optimum wavelengths with HS image features, and optimum wavelengths with GLCM image features, the model integrating optimum wavelengths with GLGCM gave the highest classification rate of 93.14% and 90.91% for calibration and validation sets, respectively. Results indicated that the classification accuracy can be improved by combining spectral features with textural features and the fusion of critical spectral and textural features had better potential than single spectral extraction in classifying fresh and frozen-thawed pork meat. Copyright © 2014 Elsevier Ltd. All rights reserved.

Developing collaborative classifiers using an expert-based model

USGS Publications Warehouse

Mountrakis, G.; Watts, R.; Luo, L.; Wang, Jingyuan

2009-01-01

This paper presents a hierarchical, multi-stage adaptive strategy for image classification. We iteratively apply various classification methods (e.g., decision trees, neural networks), identify regions of parametric and geographic space where accuracy is low, and in these regions, test and apply alternate methods repeating the process until the entire image is classified. Currently, classifiers are evaluated through human input using an expert-based system; therefore, this paper acts as the proof of concept for collaborative classifiers. Because we decompose the problem into smaller, more manageable sub-tasks, our classification exhibits increased flexibility compared to existing methods since classification methods are tailored to the idiosyncrasies of specific regions. A major benefit of our approach is its scalability and collaborative support since selected low-accuracy classifiers can be easily replaced with others without affecting classification accuracy in high accuracy areas. At each stage, we develop spatially explicit accuracy metrics that provide straightforward assessment of results by non-experts and point to areas that need algorithmic improvement or ancillary data. Our approach is demonstrated in the task of detecting impervious surface areas, an important indicator for human-induced alterations to the environment, using a 2001 Landsat scene from Las Vegas, Nevada. ?? 2009 American Society for Photogrammetry and Remote Sensing.

Verification and classification bias interactions in diagnostic test accuracy studies for fine-needle aspiration biopsy.

PubMed

Schmidt, Robert L; Walker, Brandon S; Cohen, Michael B

2015-03-01

Reliable estimates of accuracy are important for any diagnostic test. Diagnostic accuracy studies are subject to unique sources of bias. Verification bias and classification bias are 2 sources of bias that commonly occur in diagnostic accuracy studies. Statistical methods are available to estimate the impact of these sources of bias when they occur alone. The impact of interactions when these types of bias occur together has not been investigated. We developed mathematical relationships to show the combined effect of verification bias and classification bias. A wide range of case scenarios were generated to assess the impact of bias components and interactions on total bias. Interactions between verification bias and classification bias caused overestimation of sensitivity and underestimation of specificity. Interactions had more effect on sensitivity than specificity. Sensitivity was overestimated by at least 7% in approximately 6% of the tested scenarios. Specificity was underestimated by at least 7% in less than 0.1% of the scenarios. Interactions between verification bias and classification bias create distortions in accuracy estimates that are greater than would be predicted from each source of bias acting independently. © 2014 American Cancer Society.

Compensatory neurofuzzy model for discrete data classification in biomedical

NASA Astrophysics Data System (ADS)

Ceylan, Rahime

2015-03-01

Biomedical data is separated to two main sections: signals and discrete data. So, studies in this area are about biomedical signal classification or biomedical discrete data classification. There are artificial intelligence models which are relevant to classification of ECG, EMG or EEG signals. In same way, in literature, many models exist for classification of discrete data taken as value of samples which can be results of blood analysis or biopsy in medical process. Each algorithm could not achieve high accuracy rate on classification of signal and discrete data. In this study, compensatory neurofuzzy network model is presented for classification of discrete data in biomedical pattern recognition area. The compensatory neurofuzzy network has a hybrid and binary classifier. In this system, the parameters of fuzzy systems are updated by backpropagation algorithm. The realized classifier model is conducted to two benchmark datasets (Wisconsin Breast Cancer dataset and Pima Indian Diabetes dataset). Experimental studies show that compensatory neurofuzzy network model achieved 96.11% accuracy rate in classification of breast cancer dataset and 69.08% accuracy rate was obtained in experiments made on diabetes dataset with only 10 iterations.

Selection of the Australian indicator region

NASA Technical Reports Server (NTRS)

Reed, C. R. (Principal Investigator)

1981-01-01

Each Australian state was examined for the availability of LANDSAT data, area, yield, and production characteristics, statistics, crop calendars, and other ancillary data. Agrophysical conditions that could influence labeling and classification accuracies were identified in connection with the highest producing states as determined from available Australian crop statistics. Based primarily on these production statistics, Western Australia and New South Wales were selected as the wheat indicator region for Australia. The general characteristics of wheat in the indicator region, with potential problems anticipated for proportion estimation are considered. The varieties of wheat, the diseases and pests common to New South Wales, and the wheat growing regions of both states are examined.

Diagnostic accuracy of imaging devices in glaucoma: A meta-analysis.

PubMed

Fallon, Monica; Valero, Oliver; Pazos, Marta; Antón, Alfonso

Imaging devices such as the Heidelberg retinal tomograph-3 (HRT3), scanning laser polarimetry (GDx), and optical coherence tomography (OCT) play an important role in glaucoma diagnosis. A systematic search for evidence-based data was performed for prospective studies evaluating the diagnostic accuracy of HRT3, GDx, and OCT. The diagnostic odds ratio (DOR) was calculated. To compare the accuracy among instruments and parameters, a meta-analysis considering the hierarchical summary receiver-operating characteristic model was performed. The risk of bias was assessed using quality assessment of diagnostic accuracy studies, version 2. Studies in the context of screening programs were used for qualitative analysis. Eighty-six articles were included. The DOR values were 29.5 for OCT, 18.6 for GDx, and 13.9 for HRT. The heterogeneity analysis demonstrated statistically a significant influence of degree of damage and ethnicity. Studies analyzing patients with earlier glaucoma showed poorer results. The risk of bias was high for patient selection. Screening studies showed lower sensitivity values and similar specificity values when compared with those included in the meta-analysis. The classification capabilities of GDx, HRT, and OCT were high and similar across the 3 instruments. The highest estimated DOR was obtained with OCT. Diagnostic accuracy could be overestimated in studies including prediagnosed groups of subjects. Copyright © 2017 Elsevier Inc. All rights reserved.

Forest Classification Accuracy as Influenced by Multispectral Scanner Spatial Resolution. [Sam Houston National Forest, Texas

NASA Technical Reports Server (NTRS)

Nalepka, R. F. (Principal Investigator); Sadowski, F. E.; Sarno, J. E.

1976-01-01

The author has identified the following significant results. A supervised classification within two separate ground areas of the Sam Houston National Forest was carried out for two sq meters spatial resolution MSS data. Data were progressively coarsened to simulate five additional cases of spatial resolution ranging up to 64 sq meters. Similar processing and analysis of all spatial resolutions enabled evaluations of the effect of spatial resolution on classification accuracy for various levels of detail and the effects on area proportion estimation for very general forest features. For very coarse resolutions, a subset of spectral channels which simulated the proposed thematic mapper channels was used to study classification accuracy.

The use of Landsat data to inventory cotton and soybean acreage in North Alabama

NASA Technical Reports Server (NTRS)

Downs, S. W., Jr.; Faust, N. L.

1980-01-01

This study was performed to determine if Landsat data could be used to improve the accuracy of the estimation of cotton acreage. A linear classification algorithm and a maximum likelihood algorithm were used for computer classification of the area, and the classification was compared with ground truth. The classification accuracy for some fields was greater than 90 percent; however, the overall accuracy was 71 percent for cotton and 56 percent for soybeans. The results of this research indicate that computer analysis of Landsat data has potential for improving upon the methods presently being used to determine cotton acreage; however, additional experiments and refinements are needed before the method can be used operationally.

Accuracy assessment, using stratified plurality sampling, of portions of a LANDSAT classification of the Arctic National Wildlife Refuge Coastal Plain

NASA Technical Reports Server (NTRS)

Card, Don H.; Strong, Laurence L.

1989-01-01

An application of a classification accuracy assessment procedure is described for a vegetation and land cover map prepared by digital image processing of LANDSAT multispectral scanner data. A statistical sampling procedure called Stratified Plurality Sampling was used to assess the accuracy of portions of a map of the Arctic National Wildlife Refuge coastal plain. Results are tabulated as percent correct classification overall as well as per category with associated confidence intervals. Although values of percent correct were disappointingly low for most categories, the study was useful in highlighting sources of classification error and demonstrating shortcomings of the plurality sampling method.

Object-Based Paddy Rice Mapping Using HJ-1A/B Data and Temporal Features Extracted from Time Series MODIS NDVI Data

PubMed Central

Singha, Mrinal; Wu, Bingfang; Zhang, Miao

2016-01-01

Accurate and timely mapping of paddy rice is vital for food security and environmental sustainability. This study evaluates the utility of temporal features extracted from coarse resolution data for object-based paddy rice classification of fine resolution data. The coarse resolution vegetation index data is first fused with the fine resolution data to generate the time series fine resolution data. Temporal features are extracted from the fused data and added with the multi-spectral data to improve the classification accuracy. Temporal features provided the crop growth information, while multi-spectral data provided the pattern variation of paddy rice. The achieved overall classification accuracy and kappa coefficient were 84.37% and 0.68, respectively. The results indicate that the use of temporal features improved the overall classification accuracy of a single-date multi-spectral image by 18.75% from 65.62% to 84.37%. The minimum sensitivity (MS) of the paddy rice classification has also been improved. The comparison showed that the mapped paddy area was analogous to the agricultural statistics at the district level. This work also highlighted the importance of feature selection to achieve higher classification accuracies. These results demonstrate the potential of the combined use of temporal and spectral features for accurate paddy rice classification. PMID:28025525

Combining various types of classifiers and features extracted from magnetic resonance imaging data in schizophrenia recognition.

PubMed

Janousova, Eva; Schwarz, Daniel; Kasparek, Tomas

2015-06-30

We investigated a combination of three classification algorithms, namely the modified maximum uncertainty linear discriminant analysis (mMLDA), the centroid method, and the average linkage, with three types of features extracted from three-dimensional T1-weighted magnetic resonance (MR) brain images, specifically MR intensities, grey matter densities, and local deformations for distinguishing 49 first episode schizophrenia male patients from 49 healthy male subjects. The feature sets were reduced using intersubject principal component analysis before classification. By combining the classifiers, we were able to obtain slightly improved results when compared with single classifiers. The best classification performance (81.6% accuracy, 75.5% sensitivity, and 87.8% specificity) was significantly better than classification by chance. We also showed that classifiers based on features calculated using more computation-intensive image preprocessing perform better; mMLDA with classification boundary calculated as weighted mean discriminative scores of the groups had improved sensitivity but similar accuracy compared to the original MLDA; reducing a number of eigenvectors during data reduction did not always lead to higher classification accuracy, since noise as well as the signal important for classification were removed. Our findings provide important information for schizophrenia research and may improve accuracy of computer-aided diagnostics of neuropsychiatric diseases. Copyright © 2015 Elsevier Ireland Ltd. All rights reserved.

Object-Based Paddy Rice Mapping Using HJ-1A/B Data and Temporal Features Extracted from Time Series MODIS NDVI Data.

PubMed

Singha, Mrinal; Wu, Bingfang; Zhang, Miao

2016-12-22

Accurate and timely mapping of paddy rice is vital for food security and environmental sustainability. This study evaluates the utility of temporal features extracted from coarse resolution data for object-based paddy rice classification of fine resolution data. The coarse resolution vegetation index data is first fused with the fine resolution data to generate the time series fine resolution data. Temporal features are extracted from the fused data and added with the multi-spectral data to improve the classification accuracy. Temporal features provided the crop growth information, while multi-spectral data provided the pattern variation of paddy rice. The achieved overall classification accuracy and kappa coefficient were 84.37% and 0.68, respectively. The results indicate that the use of temporal features improved the overall classification accuracy of a single-date multi-spectral image by 18.75% from 65.62% to 84.37%. The minimum sensitivity (MS) of the paddy rice classification has also been improved. The comparison showed that the mapped paddy area was analogous to the agricultural statistics at the district level. This work also highlighted the importance of feature selection to achieve higher classification accuracies. These results demonstrate the potential of the combined use of temporal and spectral features for accurate paddy rice classification.

Evaluation of seabed mapping methods for fine-scale classification of extremely shallow benthic habitats - Application to the Venice Lagoon, Italy

NASA Astrophysics Data System (ADS)

Montereale Gavazzi, G.; Madricardo, F.; Janowski, L.; Kruss, A.; Blondel, P.; Sigovini, M.; Foglini, F.

2016-03-01

Recent technological developments of multibeam echosounder systems (MBES) allow mapping of benthic habitats with unprecedented detail. MBES can now be employed in extremely shallow waters, challenging data acquisition (as these instruments were often designed for deeper waters) and data interpretation (honed on datasets with resolution sometimes orders of magnitude lower). With extremely high-resolution bathymetry and co-located backscatter data, it is now possible to map the spatial distribution of fine scale benthic habitats, even identifying the acoustic signatures of single sponges. In this context, it is necessary to understand which of the commonly used segmentation methods is best suited to account for such level of detail. At the same time, new sampling protocols for precisely geo-referenced ground truth data need to be developed to validate the benthic environmental classification. This study focuses on a dataset collected in a shallow (2-10 m deep) tidal channel of the Lagoon of Venice, Italy. Using 0.05-m and 0.2-m raster grids, we compared a range of classifications, both pixel-based and object-based approaches, including manual, Maximum Likelihood Classifier, Jenks Optimization clustering, textural analysis and Object Based Image Analysis. Through a comprehensive and accurately geo-referenced ground truth dataset, we were able to identify five different classes of the substrate composition, including sponges, mixed submerged aquatic vegetation, mixed detritic bottom (fine and coarse) and unconsolidated bare sediment. We computed estimates of accuracy (namely Overall, User, Producer Accuracies and the Kappa statistic) by cross tabulating predicted and reference instances. Overall, pixel based segmentations produced the highest accuracies and the accuracy assessment is strongly dependent on the number of classes chosen for the thematic output. Tidal channels in the Venice Lagoon are extremely important in terms of habitats and sediment distribution, particularly within the context of the new tidal barrier being built. However, they had remained largely unexplored until now, because of the surveying challenges. The application of this remote sensing approach, combined with targeted sampling, opens a new perspective in the monitoring of benthic habitats in view of a knowledge-based management of natural resources in shallow coastal areas.

Aircraft-Based Satellite Navigation Augmentation to Enable Automated Landing and Movement on the Airport Surface

NASA Astrophysics Data System (ADS)

Obeidat, Qasem Turki

A brain-computer interface (BCI) enables a paralyzed user to interact with an external device through brain signals. A BCI measures identifies patterns within these measured signals, translating such patterns into commands. The P300 is a pattern of a scalp potentials elicited by a luminance increment of an attended target rather than a non-target character of an alphanumeric matrix. The Row-Column Paradigm (RCP) can utilize responses to series of illuminations of matrix target and non-target characters to spell out alphanumeric strings of P300-eliciting target characters, yet this popular RCP speller faces three challenges. Theadjacent problem concerns the proximity of neighboring characters, the crowding problem concerns their number. Both adjacent and crowding problems concern how these factors impede BCI performance. The fatigue problem concerns how RCP use is tiring. This dissertation addressed these challenges for both desktop and mobile platforms. A new P300 speller interface, the Zigzag Paradigm (ZP), reduced the adjacent problem by increasing the distance between adjacent characters, as well as the crowding problem, by reducing the number neighboring characters. In desktop study, the classification accuracy was significantly improved 91% with the ZP VS 80.6% with the RCP. Since the ZP is not suitable for mobile P300 spellers with a small screen size, a new P300 speller interface was developed in this study, the Edges Paradigm (EP). The EP reduced the adjacent and crowding problems by adding flashing squares located upon the outer edges of the character matrix in the EP. The classification accuracy of the EP (i.e., 93.3%) was significantly higher than the RCP (i.e., 82.1%). We further compared three speller paradigms (i.e., RCP, ZP, and EP), and the result indicated that the EP produced the highest accuracy and caused less fatigue. Later, the EP is implemented in a simulator of a Samsung galaxy smart phone on the Microsoft Surface Pro 2. The mobile EP was compared with the RCP under the mobility situation when a user is moving on a wheelchair. The results showed that the EP significantly improved the online classification accuracy and user experience over the RCP.

Accurate multimodal probabilistic prediction of conversion to Alzheimer's disease in patients with mild cognitive impairment.

PubMed

Young, Jonathan; Modat, Marc; Cardoso, Manuel J; Mendelson, Alex; Cash, Dave; Ourselin, Sebastien

2013-01-01

Accurately identifying the patients that have mild cognitive impairment (MCI) who will go on to develop Alzheimer's disease (AD) will become essential as new treatments will require identification of AD patients at earlier stages in the disease process. Most previous work in this area has centred around the same automated techniques used to diagnose AD patients from healthy controls, by coupling high dimensional brain image data or other relevant biomarker data to modern machine learning techniques. Such studies can now distinguish between AD patients and controls as accurately as an experienced clinician. Models trained on patients with AD and control subjects can also distinguish between MCI patients that will convert to AD within a given timeframe (MCI-c) and those that remain stable (MCI-s), although differences between these groups are smaller and thus, the corresponding accuracy is lower. The most common type of classifier used in these studies is the support vector machine, which gives categorical class decisions. In this paper, we introduce Gaussian process (GP) classification to the problem. This fully Bayesian method produces naturally probabilistic predictions, which we show correlate well with the actual chances of converting to AD within 3 years in a population of 96 MCI-s and 47 MCI-c subjects. Furthermore, we show that GPs can integrate multimodal data (in this study volumetric MRI, FDG-PET, cerebrospinal fluid, and APOE genotype with the classification process through the use of a mixed kernel). The GP approach aids combination of different data sources by learning parameters automatically from training data via type-II maximum likelihood, which we compare to a more conventional method based on cross validation and an SVM classifier. When the resulting probabilities from the GP are dichotomised to produce a binary classification, the results for predicting MCI conversion based on the combination of all three types of data show a balanced accuracy of 74%. This is a substantially higher accuracy than could be obtained using any individual modality or using a multikernel SVM, and is competitive with the highest accuracy yet achieved for predicting conversion within three years on the widely used ADNI dataset.

Automatic classification of protein structures using physicochemical parameters.

PubMed

Mohan, Abhilash; Rao, M Divya; Sunderrajan, Shruthi; Pennathur, Gautam

2014-09-01

Protein classification is the first step to functional annotation; SCOP and Pfam databases are currently the most relevant protein classification schemes. However, the disproportion in the number of three dimensional (3D) protein structures generated versus their classification into relevant superfamilies/families emphasizes the need for automated classification schemes. Predicting function of novel proteins based on sequence information alone has proven to be a major challenge. The present study focuses on the use of physicochemical parameters in conjunction with machine learning algorithms (Naive Bayes, Decision Trees, Random Forest and Support Vector Machines) to classify proteins into their respective SCOP superfamily/Pfam family, using sequence derived information. Spectrophores™, a 1D descriptor of the 3D molecular field surrounding a structure was used as a benchmark to compare the performance of the physicochemical parameters. The machine learning algorithms were modified to select features based on information gain for each SCOP superfamily/Pfam family. The effect of combining physicochemical parameters and spectrophores on classification accuracy (CA) was studied. Machine learning algorithms trained with the physicochemical parameters consistently classified SCOP superfamilies and Pfam families with a classification accuracy above 90%, while spectrophores performed with a CA of around 85%. Feature selection improved classification accuracy for both physicochemical parameters and spectrophores based machine learning algorithms. Combining both attributes resulted in a marginal loss of performance. Physicochemical parameters were able to classify proteins from both schemes with classification accuracy ranging from 90-96%. These results suggest the usefulness of this method in classifying proteins from amino acid sequences.

Classification accuracy for stratification with remotely sensed data

Treesearch

Raymond L. Czaplewski; Paul L. Patterson

2003-01-01

Tools are developed that help specify the classification accuracy required from remotely sensed data. These tools are applied during the planning stage of a sample survey that will use poststratification, prestratification with proportional allocation, or double sampling for stratification. Accuracy standards are developed in terms of an “error matrix,” which is...

Metric learning for automatic sleep stage classification.

PubMed

Phan, Huy; Do, Quan; Do, The-Luan; Vu, Duc-Lung

2013-01-01

We introduce in this paper a metric learning approach for automatic sleep stage classification based on single-channel EEG data. We show that learning a global metric from training data instead of using the default Euclidean metric, the k-nearest neighbor classification rule outperforms state-of-the-art methods on Sleep-EDF dataset with various classification settings. The overall accuracy for Awake/Sleep and 4-class classification setting are 98.32% and 94.49% respectively. Furthermore, the superior accuracy is achieved by performing classification on a low-dimensional feature space derived from time and frequency domains and without the need for artifact removal as a preprocessing step.

A Response to an Article Published in "Educational Research"'s Special Issue on Assessment (June 2009). What Can Be Inferred about Classification Accuracy from Classification Consistency?

ERIC Educational Resources Information Center

Bramley, Tom

2010-01-01

Background: A recent article published in "Educational Research" on the reliability of results in National Curriculum testing in England (Newton, "The reliability of results from national curriculum testing in England," "Educational Research" 51, no. 2: 181-212, 2009) suggested that: (1) classification accuracy can be…

Comparative Approach of MRI-Based Brain Tumor Segmentation and Classification Using Genetic Algorithm.

PubMed

Bahadure, Nilesh Bhaskarrao; Ray, Arun Kumar; Thethi, Har Pal

2018-01-17

The detection of a brain tumor and its classification from modern imaging modalities is a primary concern, but a time-consuming and tedious work was performed by radiologists or clinical supervisors. The accuracy of detection and classification of tumor stages performed by radiologists is depended on their experience only, so the computer-aided technology is very important to aid with the diagnosis accuracy. In this study, to improve the performance of tumor detection, we investigated comparative approach of different segmentation techniques and selected the best one by comparing their segmentation score. Further, to improve the classification accuracy, the genetic algorithm is employed for the automatic classification of tumor stage. The decision of classification stage is supported by extracting relevant features and area calculation. The experimental results of proposed technique are evaluated and validated for performance and quality analysis on magnetic resonance brain images, based on segmentation score, accuracy, sensitivity, specificity, and dice similarity index coefficient. The experimental results achieved 92.03% accuracy, 91.42% specificity, 92.36% sensitivity, and an average segmentation score between 0.82 and 0.93 demonstrating the effectiveness of the proposed technique for identifying normal and abnormal tissues from brain MR images. The experimental results also obtained an average of 93.79% dice similarity index coefficient, which indicates better overlap between the automated extracted tumor regions with manually extracted tumor region by radiologists.

Thematic accuracy of the National Land Cover Database (NLCD) 2001 land cover for Alaska

USGS Publications Warehouse

Selkowitz, D.J.; Stehman, S.V.

2011-01-01

The National Land Cover Database (NLCD) 2001 Alaska land cover classification is the first 30-m resolution land cover product available covering the entire state of Alaska. The accuracy assessment of the NLCD 2001 Alaska land cover classification employed a geographically stratified three-stage sampling design to select the reference sample of pixels. Reference land cover class labels were determined via fixed wing aircraft, as the high resolution imagery used for determining the reference land cover classification in the conterminous U.S. was not available for most of Alaska. Overall thematic accuracy for the Alaska NLCD was 76.2% (s.e. 2.8%) at Level II (12 classes evaluated) and 83.9% (s.e. 2.1%) at Level I (6 classes evaluated) when agreement was defined as a match between the map class and either the primary or alternate reference class label. When agreement was defined as a match between the map class and primary reference label only, overall accuracy was 59.4% at Level II and 69.3% at Level I. The majority of classification errors occurred at Level I of the classification hierarchy (i.e., misclassifications were generally to a different Level I class, not to a Level II class within the same Level I class). Classification accuracy was higher for more abundant land cover classes and for pixels located in the interior of homogeneous land cover patches. ?? 2011.

Classification accuracy on the family planning participation status using kernel discriminant analysis

NASA Astrophysics Data System (ADS)

Kurniawan, Dian; Suparti; Sugito

2018-05-01

Population growth in Indonesia has increased every year. According to the population census conducted by the Central Bureau of Statistics (BPS) in 2010, the population of Indonesia has reached 237.6 million people. Therefore, to control the population growth rate, the government hold Family Planning or Keluarga Berencana (KB) program for couples of childbearing age. The purpose of this program is to improve the health of mothers and children in order to manifest prosperous society by controlling births while ensuring control of population growth. The data used in this study is the updated family data of Semarang city in 2016 that conducted by National Family Planning Coordinating Board (BKKBN). From these data, classifiers with kernel discriminant analysis will be obtained, and also classification accuracy will be obtained from that method. The result of the analysis showed that normal kernel discriminant analysis gives 71.05 % classification accuracy with 28.95 % classification error. Whereas triweight kernel discriminant analysis gives 73.68 % classification accuracy with 26.32 % classification error. Using triweight kernel discriminant for data preprocessing of family planning participation of childbearing age couples in Semarang City of 2016 can be stated better than with normal kernel discriminant.

The Effects of Q-Matrix Design on Classification Accuracy in the Log-Linear Cognitive Diagnosis Model.

PubMed

Madison, Matthew J; Bradshaw, Laine P

2015-06-01

Diagnostic classification models are psychometric models that aim to classify examinees according to their mastery or non-mastery of specified latent characteristics. These models are well-suited for providing diagnostic feedback on educational assessments because of their practical efficiency and increased reliability when compared with other multidimensional measurement models. A priori specifications of which latent characteristics or attributes are measured by each item are a core element of the diagnostic assessment design. This item-attribute alignment, expressed in a Q-matrix, precedes and supports any inference resulting from the application of the diagnostic classification model. This study investigates the effects of Q-matrix design on classification accuracy for the log-linear cognitive diagnosis model. Results indicate that classification accuracy, reliability, and convergence rates improve when the Q-matrix contains isolated information from each measured attribute.

An improved strategy for skin lesion detection and classification using uniform segmentation and feature selection based approach.

PubMed

Nasir, Muhammad; Attique Khan, Muhammad; Sharif, Muhammad; Lali, Ikram Ullah; Saba, Tanzila; Iqbal, Tassawar

2018-02-21

Melanoma is the deadliest type of skin cancer with highest mortality rate. However, the annihilation in early stage implies a high survival rate therefore, it demands early diagnosis. The accustomed diagnosis methods are costly and cumbersome due to the involvement of experienced experts as well as the requirements for highly equipped environment. The recent advancements in computerized solutions for these diagnoses are highly promising with improved accuracy and efficiency. In this article, we proposed a method for the classification of melanoma and benign skin lesions. Our approach integrates preprocessing, lesion segmentation, features extraction, features selection, and classification. Preprocessing is executed in the context of hair removal by DullRazor, whereas lesion texture and color information are utilized to enhance the lesion contrast. In lesion segmentation, a hybrid technique has been implemented and results are fused using additive law of probability. Serial based method is applied subsequently that extracts and fuses the traits such as color, texture, and HOG (shape). The fused features are selected afterwards by implementing a novel Boltzman Entropy method. Finally, the selected features are classified by Support Vector Machine. The proposed method is evaluated on publically available data set PH2. Our approach has provided promising results of sensitivity 97.7%, specificity 96.7%, accuracy 97.5%, and F-score 97.5%, which are significantly better than the results of existing methods available on the same data set. The proposed method detects and classifies melanoma significantly good as compared to existing methods. © 2018 Wiley Periodicals, Inc.

Classification of micro-Doppler signatures of human aquatic activity through simulation and measurement using transferred learning

NASA Astrophysics Data System (ADS)

Kim, Youngwook; Park, Jinhee; Moon, Taesup

2017-05-01

Remote detection of human aquatic activity can be applied not only to ocean surveillance but also to rescue operations. When a human is illuminated by electromagnetic waves, a Doppler signal is generated from his or her moving parts. Indeed, bodily movements are what make humans' micro-Doppler signatures unique, offering a chance to classify human motions. Certain studies have analyzed and attempted to recognize human aquatic activity, but the topic has yet to be extensively studied. In the present research, we simulate the micro-Doppler signatures of a swimming person in an attempt to investigate those signatures' characteristics. We model human arms as point scatterers while assuming a simple arm motion. By means of such a simulation, we can obtain spectrograms from a swimming person, then extend our measurement to multiple participants. Measurements are taken from five aquatic activities featuring five participants, comprising freestyle, backstroke, and breaststroke, pulling a boat, and rowing. As suggested by the simulation study, the spectrograms for the five activities show different micro-Doppler signatures; hence, we propose to classify them using a deep convolutional neural network (DCNN). In particular, we suggest the use of a transfer-learned DCNN, which is based on a DCNN pretrained by a large-scale RGB image dataset that is, ImageNet. The classification accuracy is calculated using fivefold cross-validation on our dataset. We find that a DCNN trained through transfer learning achieves the highest accuracy while also providing a significant performance boost over the conventional classification method.

A wavelet-based technique to predict treatment outcome for Major Depressive Disorder.

PubMed

Mumtaz, Wajid; Xia, Likun; Mohd Yasin, Mohd Azhar; Azhar Ali, Syed Saad; Malik, Aamir Saeed

2017-01-01

Treatment management for Major Depressive Disorder (MDD) has been challenging. However, electroencephalogram (EEG)-based predictions of antidepressant's treatment outcome may help during antidepressant's selection and ultimately improve the quality of life for MDD patients. In this study, a machine learning (ML) method involving pretreatment EEG data was proposed to perform such predictions for Selective Serotonin Reuptake Inhibitor (SSRIs). For this purpose, the acquisition of experimental data involved 34 MDD patients and 30 healthy controls. Consequently, a feature matrix was constructed involving time-frequency decomposition of EEG data based on wavelet transform (WT) analysis, termed as EEG data matrix. However, the resultant EEG data matrix had high dimensionality. Therefore, dimension reduction was performed based on a rank-based feature selection method according to a criterion, i.e., receiver operating characteristic (ROC). As a result, the most significant features were identified and further be utilized during the training and testing of a classification model, i.e., the logistic regression (LR) classifier. Finally, the LR model was validated with 100 iterations of 10-fold cross-validation (10-CV). The classification results were compared with short-time Fourier transform (STFT) analysis, and empirical mode decompositions (EMD). The wavelet features extracted from frontal and temporal EEG data were found statistically significant. In comparison with other time-frequency approaches such as the STFT and EMD, the WT analysis has shown highest classification accuracy, i.e., accuracy = 87.5%, sensitivity = 95%, and specificity = 80%. In conclusion, significant wavelet coefficients extracted from frontal and temporal pre-treatment EEG data involving delta and theta frequency bands may predict antidepressant's treatment outcome for the MDD patients.

The study of vehicle classification equipment with solutions to improve accuracy in Oklahoma.

DOT National Transportation Integrated Search

2014-12-01

The accuracy of vehicle counting and classification data is vital for appropriate future highway and road : design, including determining pavement characteristics, eliminating traffic jams, and improving safety. : Organizations relying on vehicle cla...

Prediction of cause of death from forensic autopsy reports using text classification techniques: A comparative study.

PubMed

Mujtaba, Ghulam; Shuib, Liyana; Raj, Ram Gopal; Rajandram, Retnagowri; Shaikh, Khairunisa

2018-07-01

Automatic text classification techniques are useful for classifying plaintext medical documents. This study aims to automatically predict the cause of death from free text forensic autopsy reports by comparing various schemes for feature extraction, term weighing or feature value representation, text classification, and feature reduction. For experiments, the autopsy reports belonging to eight different causes of death were collected, preprocessed and converted into 43 master feature vectors using various schemes for feature extraction, representation, and reduction. The six different text classification techniques were applied on these 43 master feature vectors to construct a classification model that can predict the cause of death. Finally, classification model performance was evaluated using four performance measures i.e. overall accuracy, macro precision, macro-F-measure, and macro recall. From experiments, it was found that that unigram features obtained the highest performance compared to bigram, trigram, and hybrid-gram features. Furthermore, in feature representation schemes, term frequency, and term frequency with inverse document frequency obtained similar and better results when compared with binary frequency, and normalized term frequency with inverse document frequency. Furthermore, the chi-square feature reduction approach outperformed Pearson correlation, and information gain approaches. Finally, in text classification algorithms, support vector machine classifier outperforms random forest, Naive Bayes, k-nearest neighbor, decision tree, and ensemble-voted classifier. Our results and comparisons hold practical importance and serve as references for future works. Moreover, the comparison outputs will act as state-of-art techniques to compare future proposals with existing automated text classification techniques. Copyright © 2017 Elsevier Ltd and Faculty of Forensic and Legal Medicine. All rights reserved.

Improved supervised classification of accelerometry data to distinguish behaviors of soaring birds.

PubMed

Sur, Maitreyi; Suffredini, Tony; Wessells, Stephen M; Bloom, Peter H; Lanzone, Michael; Blackshire, Sheldon; Sridhar, Srisarguru; Katzner, Todd

2017-01-01

Soaring birds can balance the energetic costs of movement by switching between flapping, soaring and gliding flight. Accelerometers can allow quantification of flight behavior and thus a context to interpret these energetic costs. However, models to interpret accelerometry data are still being developed, rarely trained with supervised datasets, and difficult to apply. We collected accelerometry data at 140Hz from a trained golden eagle (Aquila chrysaetos) whose flight we recorded with video that we used to characterize behavior. We applied two forms of supervised classifications, random forest (RF) models and K-nearest neighbor (KNN) models. The KNN model was substantially easier to implement than the RF approach but both were highly accurate in classifying basic behaviors such as flapping (85.5% and 83.6% accurate, respectively), soaring (92.8% and 87.6%) and sitting (84.1% and 88.9%) with overall accuracies of 86.6% and 92.3% respectively. More detailed classification schemes, with specific behaviors such as banking and straight flights were well classified only by the KNN model (91.24% accurate; RF = 61.64% accurate). The RF model maintained its accuracy of classifying basic behavior classification accuracy of basic behaviors at sampling frequencies as low as 10Hz, the KNN at sampling frequencies as low as 20Hz. Classification of accelerometer data collected from free ranging birds demonstrated a strong dependence of predicted behavior on the type of classification model used. Our analyses demonstrate the consequence of different approaches to classification of accelerometry data, the potential to optimize classification algorithms with validated flight behaviors to improve classification accuracy, ideal sampling frequencies for different classification algorithms, and a number of ways to improve commonly used analytical techniques and best practices for classification of accelerometry data.

Improved supervised classification of accelerometry data to distinguish behaviors of soaring birds

PubMed Central

Suffredini, Tony; Wessells, Stephen M.; Bloom, Peter H.; Lanzone, Michael; Blackshire, Sheldon; Sridhar, Srisarguru; Katzner, Todd

2017-01-01

Soaring birds can balance the energetic costs of movement by switching between flapping, soaring and gliding flight. Accelerometers can allow quantification of flight behavior and thus a context to interpret these energetic costs. However, models to interpret accelerometry data are still being developed, rarely trained with supervised datasets, and difficult to apply. We collected accelerometry data at 140Hz from a trained golden eagle (Aquila chrysaetos) whose flight we recorded with video that we used to characterize behavior. We applied two forms of supervised classifications, random forest (RF) models and K-nearest neighbor (KNN) models. The KNN model was substantially easier to implement than the RF approach but both were highly accurate in classifying basic behaviors such as flapping (85.5% and 83.6% accurate, respectively), soaring (92.8% and 87.6%) and sitting (84.1% and 88.9%) with overall accuracies of 86.6% and 92.3% respectively. More detailed classification schemes, with specific behaviors such as banking and straight flights were well classified only by the KNN model (91.24% accurate; RF = 61.64% accurate). The RF model maintained its accuracy of classifying basic behavior classification accuracy of basic behaviors at sampling frequencies as low as 10Hz, the KNN at sampling frequencies as low as 20Hz. Classification of accelerometer data collected from free ranging birds demonstrated a strong dependence of predicted behavior on the type of classification model used. Our analyses demonstrate the consequence of different approaches to classification of accelerometry data, the potential to optimize classification algorithms with validated flight behaviors to improve classification accuracy, ideal sampling frequencies for different classification algorithms, and a number of ways to improve commonly used analytical techniques and best practices for classification of accelerometry data. PMID:28403159

Improved supervised classification of accelerometry data to distinguish behaviors of soaring birds

USGS Publications Warehouse

Sur, Maitreyi; Suffredini, Tony; Wessells, Stephen M.; Bloom, Peter H.; Lanzone, Michael J.; Blackshire, Sheldon; Sridhar, Srisarguru; Katzner, Todd

2017-01-01

Soaring birds can balance the energetic costs of movement by switching between flapping, soaring and gliding flight. Accelerometers can allow quantification of flight behavior and thus a context to interpret these energetic costs. However, models to interpret accelerometry data are still being developed, rarely trained with supervised datasets, and difficult to apply. We collected accelerometry data at 140Hz from a trained golden eagle (Aquila chrysaetos) whose flight we recorded with video that we used to characterize behavior. We applied two forms of supervised classifications, random forest (RF) models and K-nearest neighbor (KNN) models. The KNN model was substantially easier to implement than the RF approach but both were highly accurate in classifying basic behaviors such as flapping (85.5% and 83.6% accurate, respectively), soaring (92.8% and 87.6%) and sitting (84.1% and 88.9%) with overall accuracies of 86.6% and 92.3% respectively. More detailed classification schemes, with specific behaviors such as banking and straight flights were well classified only by the KNN model (91.24% accurate; RF = 61.64% accurate). The RF model maintained its accuracy of classifying basic behavior classification accuracy of basic behaviors at sampling frequencies as low as 10Hz, the KNN at sampling frequencies as low as 20Hz. Classification of accelerometer data collected from free ranging birds demonstrated a strong dependence of predicted behavior on the type of classification model used. Our analyses demonstrate the consequence of different approaches to classification of accelerometry data, the potential to optimize classification algorithms with validated flight behaviors to improve classification accuracy, ideal sampling frequencies for different classification algorithms, and a number of ways to improve commonly used analytical techniques and best practices for classification of accelerometry data.

Salience network-based classification and prediction of symptom severity in children with autism.

PubMed

Uddin, Lucina Q; Supekar, Kaustubh; Lynch, Charles J; Khouzam, Amirah; Phillips, Jennifer; Feinstein, Carl; Ryali, Srikanth; Menon, Vinod

2013-08-01

Autism spectrum disorder (ASD) affects 1 in 88 children and is characterized by a complex phenotype, including social, communicative, and sensorimotor deficits. Autism spectrum disorder has been linked with atypical connectivity across multiple brain systems, yet the nature of these differences in young children with the disorder is not well understood. To examine connectivity of large-scale brain networks and determine whether specific networks can distinguish children with ASD from typically developing (TD) children and predict symptom severity in children with ASD. Case-control study performed at Stanford University School of Medicine of 20 children 7 to 12 years old with ASD and 20 age-, sex-, and IQ-matched TD children. Between-group differences in intrinsic functional connectivity of large-scale brain networks, performance of a classifier built to discriminate children with ASD from TD children based on specific brain networks, and correlations between brain networks and core symptoms of ASD. We observed stronger functional connectivity within several large-scale brain networks in children with ASD compared with TD children. This hyperconnectivity in ASD encompassed salience, default mode, frontotemporal, motor, and visual networks. This hyperconnectivity result was replicated in an independent cohort obtained from publicly available databases. Using maps of each individual's salience network, children with ASD could be discriminated from TD children with a classification accuracy of 78%, with 75% sensitivity and 80% specificity. The salience network showed the highest classification accuracy among all networks examined, and the blood oxygen-level dependent signal in this network predicted restricted and repetitive behavior scores. The classifier discriminated ASD from TD in the independent sample with 83% accuracy, 67% sensitivity, and 100% specificity. Salience network hyperconnectivity may be a distinguishing feature in children with ASD. Quantification of brain network connectivity is a step toward developing biomarkers for objectively identifying children with ASD.

Salience Network–Based Classification and Prediction of Symptom Severity in Children With Autism

PubMed Central

Uddin, Lucina Q.; Supekar, Kaustubh; Lynch, Charles J.; Khouzam, Amirah; Phillips, Jennifer; Feinstein, Carl; Ryali, Srikanth; Menon, Vinod

2014-01-01

IMPORTANCE Autism spectrum disorder (ASD) affects 1 in 88 children and is characterized by a complex phenotype, including social, communicative, and sensorimotor deficits. Autism spectrum disorder has been linked with atypical connectivity across multiple brain systems, yet the nature of these differences in young children with the disorder is not well understood. OBJECTIVES To examine connectivity of large-scale brain networks and determine whether specific networks can distinguish children with ASD from typically developing (TD) children and predict symptom severity in children with ASD. DESIGN, SETTING, AND PARTICIPANTS Case-control study performed at Stanford University School of Medicine of 20 children 7 to 12 years old with ASD and 20 age-, sex-, and IQ-matched TD children. MAIN OUTCOMES AND MEASURES Between-group differences in intrinsic functional connectivity of large-scale brain networks, performance of a classifier built to discriminate children with ASD from TD children based on specific brain networks, and correlations between brain networks and core symptoms of ASD. RESULTS We observed stronger functional connectivity within several large-scale brain networks in children with ASD compared with TD children. This hyperconnectivity in ASD encompassed salience, default mode, frontotemporal, motor, and visual networks. This hyperconnectivity result was replicated in an independent cohort obtained from publicly available databases. Using maps of each individual’s salience network, children with ASD could be discriminated from TD children with a classification accuracy of 78%, with 75% sensitivity and 80% specificity. The salience network showed the highest classification accuracy among all networks examined, and the blood oxygen–level dependent signal in this network predicted restricted and repetitive behavior scores. The classifier discriminated ASD from TD in the independent sample with 83% accuracy, 67% sensitivity, and 100% specificity. CONCLUSIONS AND RELEVANCE Salience network hyperconnectivity may be a distinguishing feature in children with ASD. Quantification of brain network connectivity is a step toward developing biomarkers for objectively identifying children with ASD. PMID:23803651

DOE Office of Scientific and Technical Information (OSTI.GOV)

You, D; Aryal, M; Samuels, S

Purpose: A previous study showed that large sub-volumes of tumor with low blood volume (BV) (poorly perfused) in head-and-neck (HN) cancers are significantly associated with local-regional failure (LRF) after chemoradiation therapy, and could be targeted with intensified radiation doses. This study aimed to develop an automated and scalable model to extract voxel-wise contrast-enhanced temporal features of dynamic contrastenhanced (DCE) MRI in HN cancers for predicting LRF. Methods: Our model development consists of training and testing stages. The training stage includes preprocessing of individual-voxel DCE curves from tumors for intensity normalization and temporal alignment, temporal feature extraction from the curves, featuremore » selection, and training classifiers. For feature extraction, multiresolution Haar discrete wavelet transformation is applied to each DCE curve to capture temporal contrast-enhanced features. The wavelet coefficients as feature vectors are selected. Support vector machine classifiers are trained to classify tumor voxels having either low or high BV, for which a BV threshold of 7.6% is previously established and used as ground truth. The model is tested by a new dataset. The voxel-wise DCE curves for training and testing were from 14 and 8 patients, respectively. A posterior probability map of the low BV class was created to examine the tumor sub-volume classification. Voxel-wise classification accuracy was computed to evaluate performance of the model. Results: Average classification accuracies were 87.2% for training (10-fold crossvalidation) and 82.5% for testing. The lowest and highest accuracies (patient-wise) were 68.7% and 96.4%, respectively. Posterior probability maps of the low BV class showed the sub-volumes extracted by our model similar to ones defined by the BV maps with most misclassifications occurred near the sub-volume boundaries. Conclusion: This model could be valuable to support adaptive clinical trials with further validation. The framework could be extendable and scalable to extract temporal contrastenhanced features of DCE-MRI in other tumors. We would like to acknowledge NIH for funding support: UO1 CA183848.« less

Classification Accuracy of a Wearable Activity Tracker for Assessing Sedentary Behavior and Physical Activity in 3-5-Year-Old Children.

PubMed

Byun, Wonwoo; Lee, Jung-Min; Kim, Youngwon; Brusseau, Timothy A

2018-03-26

This study examined the accuracy of the Fitbit activity tracker (FF) for quantifying sedentary behavior (SB) and varying intensities of physical activity (PA) in 3-5-year-old children. Twenty-eight healthy preschool-aged children (Girls: 46%, Mean age: 4.8 ± 1.0 years) wore the FF and were directly observed while performing a set of various unstructured and structured free-living activities from sedentary to vigorous intensity. The classification accuracy of the FF for measuring SB, light PA (LPA), moderate-to-vigorous PA (MVPA), and total PA (TPA) was examined calculating Pearson correlation coefficients (r), mean absolute percent error (MAPE), Cohen's kappa ( k ), sensitivity (Se), specificity (Sp), and area under the receiver operating curve (ROC-AUC). The classification accuracies of the FF (ROC-AUC) were 0.92, 0.63, 0.77 and 0.92 for SB, LPA, MVPA and TPA, respectively. Similarly, values of kappa, Se, Sp and percentage of correct classification were consistently high for SB and TPA, but low for LPA and MVPA. The FF demonstrated excellent classification accuracy for assessing SB and TPA, but lower accuracy for classifying LPA and MVPA. Our findings suggest that the FF should be considered as a valid instrument for assessing time spent sedentary and overall physical activity in preschool-aged children.

Ground Truth Sampling and LANDSAT Accuracy Assessment

NASA Technical Reports Server (NTRS)

Robinson, J. W.; Gunther, F. J.; Campbell, W. J.

1982-01-01

It is noted that the key factor in any accuracy assessment of remote sensing data is the method used for determining the ground truth, independent of the remote sensing data itself. The sampling and accuracy procedures developed for nuclear power plant siting study are described. The purpose of the sampling procedure was to provide data for developing supervised classifications for two study sites and for assessing the accuracy of that and the other procedures used. The purpose of the accuracy assessment was to allow the comparison of the cost and accuracy of various classification procedures as applied to various data types.

Multi-Temporal Classification and Change Detection Using Uav Images

NASA Astrophysics Data System (ADS)

Makuti, S.; Nex, F.; Yang, M. Y.

2018-05-01

In this paper different methodologies for the classification and change detection of UAV image blocks are explored. UAV is not only the cheapest platform for image acquisition but it is also the easiest platform to operate in repeated data collections over a changing area like a building construction site. Two change detection techniques have been evaluated in this study: the pre-classification and the post-classification algorithms. These methods are based on three main steps: feature extraction, classification and change detection. A set of state of the art features have been used in the tests: colour features (HSV), textural features (GLCM) and 3D geometric features. For classification purposes Conditional Random Field (CRF) has been used: the unary potential was determined using the Random Forest algorithm while the pairwise potential was defined by the fully connected CRF. In the performed tests, different feature configurations and settings have been considered to assess the performance of these methods in such challenging task. Experimental results showed that the post-classification approach outperforms the pre-classification change detection method. This was analysed using the overall accuracy, where by post classification have an accuracy of up to 62.6 % and the pre classification change detection have an accuracy of 46.5 %. These results represent a first useful indication for future works and developments.

Automated structural classification of lipids by machine learning.

PubMed

Taylor, Ryan; Miller, Ryan H; Miller, Ryan D; Porter, Michael; Dalgleish, James; Prince, John T

2015-03-01

Modern lipidomics is largely dependent upon structural ontologies because of the great diversity exhibited in the lipidome, but no automated lipid classification exists to facilitate this partitioning. The size of the putative lipidome far exceeds the number currently classified, despite a decade of work. Automated classification would benefit ongoing classification efforts by decreasing the time needed and increasing the accuracy of classification while providing classifications for mass spectral identification algorithms. We introduce a tool that automates classification into the LIPID MAPS ontology of known lipids with >95% accuracy and novel lipids with 63% accuracy. The classification is based upon simple chemical characteristics and modern machine learning algorithms. The decision trees produced are intelligible and can be used to clarify implicit assumptions about the current LIPID MAPS classification scheme. These characteristics and decision trees are made available to facilitate alternative implementations. We also discovered many hundreds of lipids that are currently misclassified in the LIPID MAPS database, strongly underscoring the need for automated classification. Source code and chemical characteristic lists as SMARTS search strings are available under an open-source license at https://www.github.com/princelab/lipid_classifier. © The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.

Comparison of diagnostic classification systems for delirium with new research criteria that incorporate the three core domains.

PubMed

Trzepacz, Paula T; Meagher, David J; Franco, José G

2016-05-01

Diagnostic classification systems do not incorporate phenomenological research findings about the three core symptom domains of delirium (Attentional/Cognitive, Circadian, Higher Level Thinking). We evaluated classification performances of novel Trzepacz, Meagher, and Franco research diagnostic criteria (TMF) that incorporate those domains and ICD-10, DSM-III-R, DSM-IV, and DSM-5. Primary data analysis of 641 patients with mixed neuropsychiatric profiles. Delirium (n=429) and nondelirium (n=212) reference standard groups were identified using cluster analysis of symptoms assessed using the Delirium Rating Scale-Revised-98. Accuracy, sensitivity, specificity, positive and negative predictive values (PPV, NPV), and likelihood ratios (LR+, LR-) are reported. TMF criteria had high sensitivity and specificity (87.4% and 89.2%), more balanced than DSM-III-R (100% and 31.6%), DSM-IV (97.7% and 74.1%), DSM-5 (97.7% and 72.6%), and ICD-10 (66.2% and 100%). PPV of DSM-III-R, DSM-IV, and DSM-5 were <90.0%, while PPV for ICD-10 and TMF were >90%. ICD-10 had the lowest NPV (59.4%). TMF had the highest LR+ (8.06) and DSM-III-R the lowest LR- (0.0). Overall, values for DSM-IV and DSM-5 were similar, whereas for ICD-10 and DSM-III-R were inverse of each other. In the pre-existing cognitive impairment/dementia subsample (n=128), TMF retained its highest LR+ though specificity (58.3%) became less well balanced with sensitivity (87.9%), which still exceeded that of DSM. TMF research diagnostic criteria performed well, with more balanced sensitivity and specificity and the highest likelihood ratio for delirium identification. Reflecting the three core domains of delirium, TMF criteria may have advantages in biological research where delineation of this syndrome is important. Copyright © 2016. Published by Elsevier Inc.

An Initial Analysis of LANDSAT-4 Thematic Mapper Data for the Discrimination of Agricultural, Forested Wetland, and Urban Land Covers

NASA Technical Reports Server (NTRS)

Quattrochi, D. A.

1984-01-01

An initial analysis of LANDSAT 4 Thematic Mapper (TM) data for the discrimination of agricultural, forested wetland, and urban land covers is conducted using a scene of data collected over Arkansas and Tennessee. A classification of agricultural lands derived from multitemporal LANDSAT Multispectral Scanner (MSS) data is compared with a classification of TM data for the same area. Results from this comparative analysis show that the multitemporal MSS classification produced an overall accuracy of 80.91% while the TM classification yields an overall classification accuracy of 97.06% correct.

Testing the Potential of Vegetation Indices for Land Use/cover Classification Using High Resolution Data

NASA Astrophysics Data System (ADS)

Karakacan Kuzucu, A.; Bektas Balcik, F.

2017-11-01

Accurate and reliable land use/land cover (LULC) information obtained by remote sensing technology is necessary in many applications such as environmental monitoring, agricultural management, urban planning, hydrological applications, soil management, vegetation condition study and suitability analysis. But this information still remains a challenge especially in heterogeneous landscapes covering urban and rural areas due to spectrally similar LULC features. In parallel with technological developments, supplementary data such as satellite-derived spectral indices have begun to be used as additional bands in classification to produce data with high accuracy. The aim of this research is to test the potential of spectral vegetation indices combination with supervised classification methods and to extract reliable LULC information from SPOT 7 multispectral imagery. The Normalized Difference Vegetation Index (NDVI), the Ratio Vegetation Index (RATIO), the Soil Adjusted Vegetation Index (SAVI) were the three vegetation indices used in this study. The classical maximum likelihood classifier (MLC) and support vector machine (SVM) algorithm were applied to classify SPOT 7 image. Catalca is selected region located in the north west of the Istanbul in Turkey, which has complex landscape covering artificial surface, forest and natural area, agricultural field, quarry/mining area, pasture/scrubland and water body. Accuracy assessment of all classified images was performed through overall accuracy and kappa coefficient. The results indicated that the incorporation of these three different vegetation indices decrease the classification accuracy for the MLC and SVM classification. In addition, the maximum likelihood classification slightly outperformed the support vector machine classification approach in both overall accuracy and kappa statistics.

The Effect of Normalization in Violence Video Classification Performance

NASA Astrophysics Data System (ADS)

Ali, Ashikin; Senan, Norhalina

2017-08-01

Basically, data pre-processing is an important part of data mining. Normalization is a pre-processing stage for any type of problem statement, especially in video classification. Challenging problems that arises in video classification is because of the heterogeneous content, large variations in video quality and complex semantic meanings of the concepts involved. Therefore, to regularize this problem, it is thoughtful to ensure normalization or basically involvement of thorough pre-processing stage aids the robustness of classification performance. This process is to scale all the numeric variables into certain range to make it more meaningful for further phases in available data mining techniques. Thus, this paper attempts to examine the effect of 2 normalization techniques namely Min-max normalization and Z-score in violence video classifications towards the performance of classification rate using Multi-layer perceptron (MLP) classifier. Using Min-Max Normalization range of [0,1] the result shows almost 98% of accuracy, meanwhile Min-Max Normalization range of [-1,1] accuracy is 59% and for Z-score the accuracy is 50%.

Automated radial basis function neural network based image classification system for diabetic retinopathy detection in retinal images

NASA Astrophysics Data System (ADS)

Anitha, J.; Vijila, C. Kezi Selva; Hemanth, D. Jude

2010-02-01

Diabetic retinopathy (DR) is a chronic eye disease for which early detection is highly essential to avoid any fatal results. Image processing of retinal images emerge as a feasible tool for this early diagnosis. Digital image processing techniques involve image classification which is a significant technique to detect the abnormality in the eye. Various automated classification systems have been developed in the recent years but most of them lack high classification accuracy. Artificial neural networks are the widely preferred artificial intelligence technique since it yields superior results in terms of classification accuracy. In this work, Radial Basis function (RBF) neural network based bi-level classification system is proposed to differentiate abnormal DR Images and normal retinal images. The results are analyzed in terms of classification accuracy, sensitivity and specificity. A comparative analysis is performed with the results of the probabilistic classifier namely Bayesian classifier to show the superior nature of neural classifier. Experimental results show promising results for the neural classifier in terms of the performance measures.

Fuzzy Classification of High Resolution Remote Sensing Scenes Using Visual Attention Features.

PubMed

Li, Linyi; Xu, Tingbao; Chen, Yun

2017-01-01

In recent years the spatial resolutions of remote sensing images have been improved greatly. However, a higher spatial resolution image does not always lead to a better result of automatic scene classification. Visual attention is an important characteristic of the human visual system, which can effectively help to classify remote sensing scenes. In this study, a novel visual attention feature extraction algorithm was proposed, which extracted visual attention features through a multiscale process. And a fuzzy classification method using visual attention features (FC-VAF) was developed to perform high resolution remote sensing scene classification. FC-VAF was evaluated by using remote sensing scenes from widely used high resolution remote sensing images, including IKONOS, QuickBird, and ZY-3 images. FC-VAF achieved more accurate classification results than the others according to the quantitative accuracy evaluation indices. We also discussed the role and impacts of different decomposition levels and different wavelets on the classification accuracy. FC-VAF improves the accuracy of high resolution scene classification and therefore advances the research of digital image analysis and the applications of high resolution remote sensing images.

Fuzzy Classification of High Resolution Remote Sensing Scenes Using Visual Attention Features

PubMed Central

Xu, Tingbao; Chen, Yun

2017-01-01

In recent years the spatial resolutions of remote sensing images have been improved greatly. However, a higher spatial resolution image does not always lead to a better result of automatic scene classification. Visual attention is an important characteristic of the human visual system, which can effectively help to classify remote sensing scenes. In this study, a novel visual attention feature extraction algorithm was proposed, which extracted visual attention features through a multiscale process. And a fuzzy classification method using visual attention features (FC-VAF) was developed to perform high resolution remote sensing scene classification. FC-VAF was evaluated by using remote sensing scenes from widely used high resolution remote sensing images, including IKONOS, QuickBird, and ZY-3 images. FC-VAF achieved more accurate classification results than the others according to the quantitative accuracy evaluation indices. We also discussed the role and impacts of different decomposition levels and different wavelets on the classification accuracy. FC-VAF improves the accuracy of high resolution scene classification and therefore advances the research of digital image analysis and the applications of high resolution remote sensing images. PMID:28761440

Protein classification based on text document classification techniques.

PubMed

Cheng, Betty Yee Man; Carbonell, Jaime G; Klein-Seetharaman, Judith

2005-03-01

The need for accurate, automated protein classification methods continues to increase as advances in biotechnology uncover new proteins. G-protein coupled receptors (GPCRs) are a particularly difficult superfamily of proteins to classify due to extreme diversity among its members. Previous comparisons of BLAST, k-nearest neighbor (k-NN), hidden markov model (HMM) and support vector machine (SVM) using alignment-based features have suggested that classifiers at the complexity of SVM are needed to attain high accuracy. Here, analogous to document classification, we applied Decision Tree and Naive Bayes classifiers with chi-square feature selection on counts of n-grams (i.e. short peptide sequences of length n) to this classification task. Using the GPCR dataset and evaluation protocol from the previous study, the Naive Bayes classifier attained an accuracy of 93.0 and 92.4% in level I and level II subfamily classification respectively, while SVM has a reported accuracy of 88.4 and 86.3%. This is a 39.7 and 44.5% reduction in residual error for level I and level II subfamily classification, respectively. The Decision Tree, while inferior to SVM, outperforms HMM in both level I and level II subfamily classification. For those GPCR families whose profiles are stored in the Protein FAMilies database of alignments and HMMs (PFAM), our method performs comparably to a search against those profiles. Finally, our method can be generalized to other protein families by applying it to the superfamily of nuclear receptors with 94.5, 97.8 and 93.6% accuracy in family, level I and level II subfamily classification respectively. Copyright 2005 Wiley-Liss, Inc.

A canonical correlation analysis based EMG classification algorithm for eliminating electrode shift effect.

PubMed

Zhe Fan; Zhong Wang; Guanglin Li; Ruomei Wang

2016-08-01

Motion classification system based on surface Electromyography (sEMG) pattern recognition has achieved good results in experimental condition. But it is still a challenge for clinical implement and practical application. Many factors contribute to the difficulty of clinical use of the EMG based dexterous control. The most obvious and important is the noise in the EMG signal caused by electrode shift, muscle fatigue, motion artifact, inherent instability of signal and biological signals such as Electrocardiogram. In this paper, a novel method based on Canonical Correlation Analysis (CCA) was developed to eliminate the reduction of classification accuracy caused by electrode shift. The average classification accuracy of our method were above 95% for the healthy subjects. In the process, we validated the influence of electrode shift on motion classification accuracy and discovered the strong correlation with correlation coefficient of >0.9 between shift position data and normal position data.

Classification of driver fatigue in an electroencephalography-based countermeasure system with source separation module.

PubMed

Rifai Chai; Naik, Ganesh R; Tran, Yvonne; Sai Ho Ling; Craig, Ashley; Nguyen, Hung T

2015-08-01

An electroencephalography (EEG)-based counter measure device could be used for fatigue detection during driving. This paper explores the classification of fatigue and alert states using power spectral density (PSD) as a feature extractor and fuzzy swarm based-artificial neural network (ANN) as a classifier. An independent component analysis of entropy rate bound minimization (ICA-ERBM) is investigated as a novel source separation technique for fatigue classification using EEG analysis. A comparison of the classification accuracy of source separator versus no source separator is presented. Classification performance based on 43 participants without the inclusion of the source separator resulted in an overall sensitivity of 71.67%, a specificity of 75.63% and an accuracy of 73.65%. However, these results were improved after the inclusion of a source separator module, resulting in an overall sensitivity of 78.16%, a specificity of 79.60% and an accuracy of 78.88% (p <; 0.05).

Forest tree species discrimination in western Himalaya using EO-1 Hyperion

NASA Astrophysics Data System (ADS)

George, Rajee; Padalia, Hitendra; Kushwaha, S. P. S.

2014-05-01

The information acquired in the narrow bands of hyperspectral remote sensing data has potential to capture plant species spectral variability, thereby improving forest tree species mapping. This study assessed the utility of spaceborne EO-1 Hyperion data in discrimination and classification of broadleaved evergreen and conifer forest tree species in western Himalaya. The pre-processing of 242 bands of Hyperion data resulted into 160 noise-free and vertical stripe corrected reflectance bands. Of these, 29 bands were selected through step-wise exclusion of bands (Wilk's Lambda). Spectral Angle Mapper (SAM) and Support Vector Machine (SVM) algorithms were applied to the selected bands to assess their effectiveness in classification. SVM was also applied to broadband data (Landsat TM) to compare the variation in classification accuracy. All commonly occurring six gregarious tree species, viz., white oak, brown oak, chir pine, blue pine, cedar and fir in western Himalaya could be effectively discriminated. SVM produced a better species classification (overall accuracy 82.27%, kappa statistic 0.79) than SAM (overall accuracy 74.68%, kappa statistic 0.70). It was noticed that classification accuracy achieved with Hyperion bands was significantly higher than Landsat TM bands (overall accuracy 69.62%, kappa statistic 0.65). Study demonstrated the potential utility of narrow spectral bands of Hyperion data in discriminating tree species in a hilly terrain.

Comparison of accuracy of fibrosis degree classifications by liver biopsy and non-invasive tests in chronic hepatitis C.

PubMed

Boursier, Jérôme; Bertrais, Sandrine; Oberti, Frédéric; Gallois, Yves; Fouchard-Hubert, Isabelle; Rousselet, Marie-Christine; Zarski, Jean-Pierre; Calès, Paul

2011-11-30

Non-invasive tests have been constructed and evaluated mainly for binary diagnoses such as significant fibrosis. Recently, detailed fibrosis classifications for several non-invasive tests have been developed, but their accuracy has not been thoroughly evaluated in comparison to liver biopsy, especially in clinical practice and for Fibroscan. Therefore, the main aim of the present study was to evaluate the accuracy of detailed fibrosis classifications available for non-invasive tests and liver biopsy. The secondary aim was to validate these accuracies in independent populations. Four HCV populations provided 2,068 patients with liver biopsy, four different pathologist skill-levels and non-invasive tests. Results were expressed as percentages of correctly classified patients. In population #1 including 205 patients and comparing liver biopsy (reference: consensus reading by two experts) and blood tests, Metavir fibrosis (FM) stage accuracy was 64.4% in local pathologists vs. 82.2% (p < 10-3) in single expert pathologist. Significant discrepancy (≥ 2FM vs reference histological result) rates were: Fibrotest: 17.2%, FibroMeter2G: 5.6%, local pathologists: 4.9%, FibroMeter3G: 0.5%, expert pathologist: 0% (p < 10-3). In population #2 including 1,056 patients and comparing blood tests, the discrepancy scores, taking into account the error magnitude, of detailed fibrosis classification were significantly different between FibroMeter2G (0.30 ± 0.55) and FibroMeter3G (0.14 ± 0.37, p < 10-3) or Fibrotest (0.84 ± 0.80, p < 10-3). In population #3 (and #4) including 458 (359) patients and comparing blood tests and Fibroscan, accuracies of detailed fibrosis classification were, respectively: Fibrotest: 42.5% (33.5%), Fibroscan: 64.9% (50.7%), FibroMeter2G: 68.7% (68.2%), FibroMeter3G: 77.1% (83.4%), p < 10-3 (p < 10-3). Significant discrepancy (≥ 2 FM) rates were, respectively: Fibrotest: 21.3% (22.2%), Fibroscan: 12.9% (12.3%), FibroMeter2G: 5.7% (6.0%), FibroMeter3G: 0.9% (0.9%), p < 10-3 (p < 10-3). The accuracy in detailed fibrosis classification of the best-performing blood test outperforms liver biopsy read by a local pathologist, i.e., in clinical practice; however, the classification precision is apparently lesser. This detailed classification accuracy is much lower than that of significant fibrosis with Fibroscan and even Fibrotest but higher with FibroMeter3G. FibroMeter classification accuracy was significantly higher than those of other non-invasive tests. Finally, for hepatitis C evaluation in clinical practice, fibrosis degree can be evaluated using an accurate blood test.

Comparison of accuracy of fibrosis degree classifications by liver biopsy and non-invasive tests in chronic hepatitis C

PubMed Central

2011-01-01

Background Non-invasive tests have been constructed and evaluated mainly for binary diagnoses such as significant fibrosis. Recently, detailed fibrosis classifications for several non-invasive tests have been developed, but their accuracy has not been thoroughly evaluated in comparison to liver biopsy, especially in clinical practice and for Fibroscan. Therefore, the main aim of the present study was to evaluate the accuracy of detailed fibrosis classifications available for non-invasive tests and liver biopsy. The secondary aim was to validate these accuracies in independent populations. Methods Four HCV populations provided 2,068 patients with liver biopsy, four different pathologist skill-levels and non-invasive tests. Results were expressed as percentages of correctly classified patients. Results In population #1 including 205 patients and comparing liver biopsy (reference: consensus reading by two experts) and blood tests, Metavir fibrosis (FM) stage accuracy was 64.4% in local pathologists vs. 82.2% (p < 10-3) in single expert pathologist. Significant discrepancy (≥ 2FM vs reference histological result) rates were: Fibrotest: 17.2%, FibroMeter2G: 5.6%, local pathologists: 4.9%, FibroMeter3G: 0.5%, expert pathologist: 0% (p < 10-3). In population #2 including 1,056 patients and comparing blood tests, the discrepancy scores, taking into account the error magnitude, of detailed fibrosis classification were significantly different between FibroMeter2G (0.30 ± 0.55) and FibroMeter3G (0.14 ± 0.37, p < 10-3) or Fibrotest (0.84 ± 0.80, p < 10-3). In population #3 (and #4) including 458 (359) patients and comparing blood tests and Fibroscan, accuracies of detailed fibrosis classification were, respectively: Fibrotest: 42.5% (33.5%), Fibroscan: 64.9% (50.7%), FibroMeter2G: 68.7% (68.2%), FibroMeter3G: 77.1% (83.4%), p < 10-3 (p < 10-3). Significant discrepancy (≥ 2 FM) rates were, respectively: Fibrotest: 21.3% (22.2%), Fibroscan: 12.9% (12.3%), FibroMeter2G: 5.7% (6.0%), FibroMeter3G: 0.9% (0.9%), p < 10-3 (p < 10-3). Conclusions The accuracy in detailed fibrosis classification of the best-performing blood test outperforms liver biopsy read by a local pathologist, i.e., in clinical practice; however, the classification precision is apparently lesser. This detailed classification accuracy is much lower than that of significant fibrosis with Fibroscan and even Fibrotest but higher with FibroMeter3G. FibroMeter classification accuracy was significantly higher than those of other non-invasive tests. Finally, for hepatitis C evaluation in clinical practice, fibrosis degree can be evaluated using an accurate blood test. PMID:22129438

The EO-1 hyperion and advanced land imager sensors for use in tundra classification studies within the Upper Kuparuk River Basin, Alaska

NASA Astrophysics Data System (ADS)

Hall-Brown, Mary

The heterogeneity of Arctic vegetation can make land cover classification vey difficult when using medium to small resolution imagery (Schneider et al., 2009; Muller et al., 1999). Using high radiometric and spatial resolution imagery, such as the SPOT 5 and IKONOS satellites, have helped arctic land cover classification accuracies rise into the 80 and 90 percentiles (Allard, 2003; Stine et al., 2010; Muller et al., 1999). However, those increases usually come at a high price. High resolution imagery is very expensive and can often add tens of thousands of dollars onto the cost of the research. The EO-1 satellite launched in 2002 carries two sensors that have high specral and/or high spatial resolutions and can be an acceptable compromise between the resolution versus cost issues. The Hyperion is a hyperspectral sensor with the capability of collecting 242 spectral bands of information. The Advanced Land Imager (ALI) is an advanced multispectral sensor whose spatial resolution can be sharpened to 10 meters. This dissertation compares the accuracies of arctic land cover classifications produced by the Hyperion and ALI sensors to the classification accuracies produced by the Systeme Pour l' Observation de le Terre (SPOT), the Landsat Thematic Mapper (TM) and the Landsat Enhanced Thematic Mapper Plus (ETM+) sensors. Hyperion and ALI images from August 2004 were collected over the Upper Kuparuk River Basin, Alaska. Image processing included the stepwise discriminant analysis of pixels that were positively classified from coinciding ground control points, geometric and radiometric correction, and principle component analysis. Finally, stratified random sampling was used to perform accuracy assessments on satellite derived land cover classifications. Accuracy was estimated from an error matrix (confusion matrix) that provided the overall, producer's and user's accuracies. This research found that while the Hyperion sensor produced classfication accuracies that were equivalent to the TM and ETM+ sensor (approximately 78%), the Hyperion could not obtain the accuracy of the SPOT 5 HRV sensor. However, the land cover classifications derived from the ALI sensor exceeded most classification accuracies derived from the TM and ETM+ senors and were even comparable to most SPOT 5 HRV classifications (87%). With the deactivation of the Landsat series satellites, the monitoring of remote locations such as in the Arctic on an uninterupted basis thoughout the world is in jeopardy. The utilization of the Hyperion and ALI sensors are a way to keep that endeavor operational. By keeping the ALI sensor active at all times, uninterupted observation of the entire Earth can be accomplished. Keeping the Hyperion sensor as a "tasked" sensor can provide scientists with additional imagery and options for their studies without overburdening storage issues.

Longitudinal evaluation of criteria for subjective cognitive decline and preclinical Alzheimer's disease in a memory clinic sample.

PubMed

Eckerström, Marie; Göthlin, Mattias; Rolstad, Sindre; Hessen, Erik; Eckerström, Carl; Nordlund, Arto; Johansson, Boo; Svensson, Johan; Jonsson, Michael; Sacuiu, Simona; Wallin, Anders

2017-01-01

Subjective cognitive decline (SCD) and biomarker-based "at-risk" concepts such as "preclinical" Alzheimer's disease (AD) have been developed to predict AD dementia before objective cognitive impairment is detectable. We longitudinally evaluated cognitive outcome when using these classifications. Memory clinic patients ( n  = 235) were classified as SCD ( n  = 122): subtle cognitive decline ( n  = 36) and mild cognitive impairment ( n  = 77) and subsequently subclassified into SCDplus and National Institute on Aging-Alzheimer's Association (NIA-AA) stages 0 to 3. Mean (standard deviation) follow-up time was 48 (35) months. Proportion declining cognitively and prognostic accuracy for cognitive decline was calculated for all classifications. Among SCDplus patients, 43% to 48% declined cognitively. Among NIA-AA stage 1 to 3 patients, 50% to 100% declined cognitively. The highest positive likelihood ratios (+LRs) for subsequent cognitive decline (+LR 6.3), dementia (+LR 3.4), and AD dementia (+LR 6.5) were found for NIA-AA stage 2. In a memory clinic setting, NIA-AA stage 2 seems to be the most successful classification in predicting objective cognitive decline, dementia, and AD dementia.

Predicting adherence of patients with HF through machine learning techniques.

PubMed

Karanasiou, Georgia Spiridon; Tripoliti, Evanthia Eleftherios; Papadopoulos, Theofilos Grigorios; Kalatzis, Fanis Georgios; Goletsis, Yorgos; Naka, Katerina Kyriakos; Bechlioulis, Aris; Errachid, Abdelhamid; Fotiadis, Dimitrios Ioannis

2016-09-01

Heart failure (HF) is a chronic disease characterised by poor quality of life, recurrent hospitalisation and high mortality. Adherence of patient to treatment suggested by the experts has been proven a significant deterrent of the above-mentioned serious consequences. However, the non-adherence rates are significantly high; a fact that highlights the importance of predicting the adherence of the patient and enabling experts to adjust accordingly patient monitoring and management. The aim of this work is to predict the adherence of patients with HF, through the application of machine learning techniques. Specifically, it aims to classify a patient not only as medication adherent or not, but also as adherent or not in terms of medication, nutrition and physical activity (global adherent). Two classification problems are addressed: (i) if the patient is global adherent or not and (ii) if the patient is medication adherent or not. About 11 classification algorithms are employed and combined with feature selection and resampling techniques. The classifiers are evaluated on a dataset of 90 patients. The patients are characterised as medication and global adherent, based on clinician estimation. The highest detection accuracy is 82 and 91% for the first and the second classification problem, respectively.

Content Abstract Classification Using Naive Bayes

NASA Astrophysics Data System (ADS)

Latif, Syukriyanto; Suwardoyo, Untung; Aldrin Wihelmus Sanadi, Edwin

2018-03-01

This study aims to classify abstract content based on the use of the highest number of words in an abstract content of the English language journals. This research uses a system of text mining technology that extracts text data to search information from a set of documents. Abstract content of 120 data downloaded at www.computer.org. Data grouping consists of three categories: DM (Data Mining), ITS (Intelligent Transport System) and MM (Multimedia). Systems built using naive bayes algorithms to classify abstract journals and feature selection processes using term weighting to give weight to each word. Dimensional reduction techniques to reduce the dimensions of word counts rarely appear in each document based on dimensional reduction test parameters of 10% -90% of 5.344 words. The performance of the classification system is tested by using the Confusion Matrix based on comparative test data and test data. The results showed that the best classification results were obtained during the 75% training data test and 25% test data from the total data. Accuracy rates for categories of DM, ITS and MM were 100%, 100%, 86%. respectively with dimension reduction parameters of 30% and the value of learning rate between 0.1-0.5.

Comparative Analysis of Haar and Daubechies Wavelet for Hyper Spectral Image Classification

NASA Astrophysics Data System (ADS)

Sharif, I.; Khare, S.

2014-11-01

With the number of channels in the hundreds instead of in the tens Hyper spectral imagery possesses much richer spectral information than multispectral imagery. The increased dimensionality of such Hyper spectral data provides a challenge to the current technique for analyzing data. Conventional classification methods may not be useful without dimension reduction pre-processing. So dimension reduction has become a significant part of Hyper spectral image processing. This paper presents a comparative analysis of the efficacy of Haar and Daubechies wavelets for dimensionality reduction in achieving image classification. Spectral data reduction using Wavelet Decomposition could be useful because it preserves the distinction among spectral signatures. Daubechies wavelets optimally capture the polynomial trends while Haar wavelet is discontinuous and resembles a step function. The performance of these wavelets are compared in terms of classification accuracy and time complexity. This paper shows that wavelet reduction has more separate classes and yields better or comparable classification accuracy. In the context of the dimensionality reduction algorithm, it is found that the performance of classification of Daubechies wavelets is better as compared to Haar wavelet while Daubechies takes more time compare to Haar wavelet. The experimental results demonstrate the classification system consistently provides over 84% classification accuracy.

ANALYSIS OF A CLASSIFICATION ERROR MATRIX USING CATEGORICAL DATA TECHNIQUES.

USGS Publications Warehouse

Rosenfield, George H.; Fitzpatrick-Lins, Katherine

1984-01-01

Summary form only given. A classification error matrix typically contains tabulation results of an accuracy evaluation of a thematic classification, such as that of a land use and land cover map. The diagonal elements of the matrix represent the counts corrected, and the usual designation of classification accuracy has been the total percent correct. The nondiagonal elements of the matrix have usually been neglected. The classification error matrix is known in statistical terms as a contingency table of categorical data. As an example, an application of these methodologies to a problem of remotely sensed data concerning two photointerpreters and four categories of classification indicated that there is no significant difference in the interpretation between the two photointerpreters, and that there are significant differences among the interpreted category classifications. However, two categories, oak and cottonwood, are not separable in classification in this experiment at the 0. 51 percent probability. A coefficient of agreement is determined for the interpreted map as a whole, and individually for each of the interpreted categories. A conditional coefficient of agreement for the individual categories is compared to other methods for expressing category accuracy which have already been presented in the remote sensing literature.

Research on Remote Sensing Image Classification Based on Feature Level Fusion

NASA Astrophysics Data System (ADS)

Yuan, L.; Zhu, G.

2018-04-01

Remote sensing image classification, as an important direction of remote sensing image processing and application, has been widely studied. However, in the process of existing classification algorithms, there still exists the phenomenon of misclassification and missing points, which leads to the final classification accuracy is not high. In this paper, we selected Sentinel-1A and Landsat8 OLI images as data sources, and propose a classification method based on feature level fusion. Compare three kind of feature level fusion algorithms (i.e., Gram-Schmidt spectral sharpening, Principal Component Analysis transform and Brovey transform), and then select the best fused image for the classification experimental. In the classification process, we choose four kinds of image classification algorithms (i.e. Minimum distance, Mahalanobis distance, Support Vector Machine and ISODATA) to do contrast experiment. We use overall classification precision and Kappa coefficient as the classification accuracy evaluation criteria, and the four classification results of fused image are analysed. The experimental results show that the fusion effect of Gram-Schmidt spectral sharpening is better than other methods. In four kinds of classification algorithms, the fused image has the best applicability to Support Vector Machine classification, the overall classification precision is 94.01 % and the Kappa coefficients is 0.91. The fused image with Sentinel-1A and Landsat8 OLI is not only have more spatial information and spectral texture characteristics, but also enhances the distinguishing features of the images. The proposed method is beneficial to improve the accuracy and stability of remote sensing image classification.

Development of classification models to detect Salmonella Enteritidis and Salmonella Typhimurium found in poultry carcass rinses by visible-near infrared hyperspectral imaging

NASA Astrophysics Data System (ADS)

Seo, Young Wook; Yoon, Seung Chul; Park, Bosoon; Hinton, Arthur; Windham, William R.; Lawrence, Kurt C.

2013-05-01

Salmonella is a major cause of foodborne disease outbreaks resulting from the consumption of contaminated food products in the United States. This paper reports the development of a hyperspectral imaging technique for detecting and differentiating two of the most common Salmonella serotypes, Salmonella Enteritidis (SE) and Salmonella Typhimurium (ST), from background microflora that are often found in poultry carcass rinse. Presumptive positive screening of colonies with a traditional direct plating method is a labor intensive and time consuming task. Thus, this paper is concerned with the detection of differences in spectral characteristics among the pure SE, ST, and background microflora grown on brilliant green sulfa (BGS) and xylose lysine tergitol 4 (XLT4) agar media with a spread plating technique. Visible near-infrared hyperspectral imaging, providing the spectral and spatial information unique to each microorganism, was utilized to differentiate SE and ST from the background microflora. A total of 10 classification models, including five machine learning algorithms, each without and with principal component analysis (PCA), were validated and compared to find the best model in classification accuracy. The five machine learning (classification) algorithms used in this study were Mahalanobis distance (MD), k-nearest neighbor (kNN), linear discriminant analysis (LDA), quadratic discriminant analysis (QDA), and support vector machine (SVM). The average classification accuracy of all 10 models on a calibration (or training) set of the pure cultures on BGS agar plates was 98% (Kappa coefficient = 0.95) in determining the presence of SE and/or ST although it was difficult to differentiate between SE and ST. The average classification accuracy of all 10 models on a training set for ST detection on XLT4 agar was over 99% (Kappa coefficient = 0.99) although SE colonies on XLT4 agar were difficult to differentiate from background microflora. The average classification accuracy of all 10 models on a validation set of chicken carcass rinses spiked with SE or ST and incubated on BGS agar plates was 94.45% and 83.73%, without and with PCA for classification, respectively. The best performing classification model on the validation set was QDA without PCA by achieving the classification accuracy of 98.65% (Kappa coefficient=0.98). The overall best performing classification model regardless of using PCA was MD with the classification accuracy of 94.84% (Kappa coefficient=0.88) on the validation set.

AVHRR composite period selection for land cover classification

USGS Publications Warehouse

Maxwell, S.K.; Hoffer, R.M.; Chapman, P.L.

2002-01-01

Multitemporal satellite image datasets provide valuable information on the phenological characteristics of vegetation, thereby significantly increasing the accuracy of cover type classifications compared to single date classifications. However, the processing of these datasets can become very complex when dealing with multitemporal data combined with multispectral data. Advanced Very High Resolution Radiometer (AVHRR) biweekly composite data are commonly used to classify land cover over large regions. Selecting a subset of these biweekly composite periods may be required to reduce the complexity and cost of land cover mapping. The objective of our research was to evaluate the effect of reducing the number of composite periods and altering the spacing of those composite periods on classification accuracy. Because inter-annual variability can have a major impact on classification results, 5 years of AVHRR data were evaluated. AVHRR biweekly composite images for spectral channels 1-4 (visible, near-infrared and two thermal bands) covering the entire growing season were used to classify 14 cover types over the entire state of Colorado for each of five different years. A supervised classification method was applied to maintain consistent procedures for each case tested. Results indicate that the number of composite periods can be halved-reduced from 14 composite dates to seven composite dates-without significantly reducing overall classification accuracy (80.4% Kappa accuracy for the 14-composite data-set as compared to 80.0% for a seven-composite dataset). At least seven composite periods were required to ensure the classification accuracy was not affected by inter-annual variability due to climate fluctuations. Concentrating more composites near the beginning and end of the growing season, as compared to using evenly spaced time periods, consistently produced slightly higher classification values over the 5 years tested (average Kappa) of 80.3% for the heavy early/late case as compared to 79.0% for the alternate dataset case).

Determination of quality television programmes based on sentiment analysis on Twitter

NASA Astrophysics Data System (ADS)

Amalia, A.; Oktinas, W.; Aulia, I.; Rahmat, R. F.

2018-03-01

Public sentiment from social media like Twitter can be used as one of the indicators to determine the quality of TV Programmes. In this study, we implemented information extraction on Twitter by using sentiment analysis method to assess the quality of TV Programmes. The first stage of this study is pre-processing which consists of cleansing, case folding, tokenizing, stop-word removal, stemming, and redundancy filtering. The next stage is weighting process for every single word by using TF-IDF method. The last step of this study is the sentiment classification process which is divided into three sentiment category which is positive, negative and neutral. We classify the TV programmes into several categories such as news, children, or films/soap operas. We implemented an improved k-nearest neighbor method in classification 4000 twitter status, for four biggest TV stations in Indonesia, with ratio 70% data for training and 30% of data for the testing process. The result obtained from this research generated the highest accuracy with k=10 as big as 90%.

Prediction of B-cell linear epitopes with a combination of support vector machine classification and amino acid propensity identification.

PubMed

Wang, Hsin-Wei; Lin, Ya-Chi; Pai, Tun-Wen; Chang, Hao-Teng

2011-01-01

Epitopes are antigenic determinants that are useful because they induce B-cell antibody production and stimulate T-cell activation. Bioinformatics can enable rapid, efficient prediction of potential epitopes. Here, we designed a novel B-cell linear epitope prediction system called LEPS, Linear Epitope Prediction by Propensities and Support Vector Machine, that combined physico-chemical propensity identification and support vector machine (SVM) classification. We tested the LEPS on four datasets: AntiJen, HIV, a newly generated PC, and AHP, a combination of these three datasets. Peptides with globally or locally high physicochemical propensities were first identified as primitive linear epitope (LE) candidates. Then, candidates were classified with the SVM based on the unique features of amino acid segments. This reduced the number of predicted epitopes and enhanced the positive prediction value (PPV). Compared to four other well-known LE prediction systems, the LEPS achieved the highest accuracy (72.52%), specificity (84.22%), PPV (32.07%), and Matthews' correlation coefficient (10.36%).

Comparison of Hybrid Classifiers for Crop Classification Using Normalized Difference Vegetation Index Time Series: A Case Study for Major Crops in North Xinjiang, China

PubMed Central

Hao, Pengyu; Wang, Li; Niu, Zheng

2015-01-01

A range of single classifiers have been proposed to classify crop types using time series vegetation indices, and hybrid classifiers are used to improve discriminatory power. Traditional fusion rules use the product of multi-single classifiers, but that strategy cannot integrate the classification output of machine learning classifiers. In this research, the performance of two hybrid strategies, multiple voting (M-voting) and probabilistic fusion (P-fusion), for crop classification using NDVI time series were tested with different training sample sizes at both pixel and object levels, and two representative counties in north Xinjiang were selected as study area. The single classifiers employed in this research included Random Forest (RF), Support Vector Machine (SVM), and See 5 (C 5.0). The results indicated that classification performance improved (increased the mean overall accuracy by 5%~10%, and reduced standard deviation of overall accuracy by around 1%) substantially with the training sample number, and when the training sample size was small (50 or 100 training samples), hybrid classifiers substantially outperformed single classifiers with higher mean overall accuracy (1%~2%). However, when abundant training samples (4,000) were employed, single classifiers could achieve good classification accuracy, and all classifiers obtained similar performances. Additionally, although object-based classification did not improve accuracy, it resulted in greater visual appeal, especially in study areas with a heterogeneous cropping pattern. PMID:26360597

Diagnostic Classification of Schizophrenia Patients on the Basis of Regional Reward-Related fMRI Signal Patterns

PubMed Central

Koch, Stefan P.; Hägele, Claudia; Haynes, John-Dylan; Heinz, Andreas; Schlagenhauf, Florian; Sterzer, Philipp

2015-01-01

Functional neuroimaging has provided evidence for altered function of mesolimbic circuits implicated in reward processing, first and foremost the ventral striatum, in patients with schizophrenia. While such findings based on significant group differences in brain activations can provide important insights into the pathomechanisms of mental disorders, the use of neuroimaging results from standard univariate statistical analysis for individual diagnosis has proven difficult. In this proof of concept study, we tested whether the predictive accuracy for the diagnostic classification of schizophrenia patients vs. healthy controls could be improved using multivariate pattern analysis (MVPA) of regional functional magnetic resonance imaging (fMRI) activation patterns for the anticipation of monetary reward. With a searchlight MVPA approach using support vector machine classification, we found that the diagnostic category could be predicted from local activation patterns in frontal, temporal, occipital and midbrain regions, with a maximal cluster peak classification accuracy of 93% for the right pallidum. Region-of-interest based MVPA for the ventral striatum achieved a maximal cluster peak accuracy of 88%, whereas the classification accuracy on the basis of standard univariate analysis reached only 75%. Moreover, using support vector regression we could additionally predict the severity of negative symptoms from ventral striatal activation patterns. These results show that MVPA can be used to substantially increase the accuracy of diagnostic classification on the basis of task-related fMRI signal patterns in a regionally specific way. PMID:25799236

Reliable classification of children’s fractures according to the comprehensive classification of long bone fractures by Müller

PubMed Central

2013-01-01

Background and purpose Guidelines for fracture treatment and evaluation require a valid classification. Classifications especially designed for children are available, but they might lead to reduced accuracy, considering the relative infrequency of childhood fractures in a general orthopedic department. We tested the reliability and accuracy of the Müller classification when used for long bone fractures in children. Methods We included all long bone fractures in children aged < 16 years who were treated in 2008 at the surgical ward of Stavanger University Hospital. 20 surgeons recorded 232 fractures. Datasets were generated for intra- and inter-rater analysis, as well as a reference dataset for accuracy calculations. We present proportion of agreement (PA) and kappa (K) statistics. Results For intra-rater analysis, overall agreement (κ) was 0.75 (95% CI: 0.68–0.81) and PA was 79%. For inter-rater assessment, K was 0.71 (95% CI: 0.61–0.80) and PA was 77%. Accuracy was estimated: κ = 0.72 (95% CI: 0.64–0.79) and PA = 76%. Interpretation The Müller classification (slightly adjusted for pediatric fractures) showed substantial to excellent accuracy among general orthopedic surgeons when applied to long bone fractures in children. However, separate knowledge about the child-specific fracture pattern, the maturity of the bone, and the degree of displacement must be considered when the treatment and the prognosis of the fractures are evaluated. PMID:23245225

Feature Selection Has a Large Impact on One-Class Classification Accuracy for MicroRNAs in Plants.

PubMed

Yousef, Malik; Saçar Demirci, Müşerref Duygu; Khalifa, Waleed; Allmer, Jens

2016-01-01

MicroRNAs (miRNAs) are short RNA sequences involved in posttranscriptional gene regulation. Their experimental analysis is complicated and, therefore, needs to be supplemented with computational miRNA detection. Currently computational miRNA detection is mainly performed using machine learning and in particular two-class classification. For machine learning, the miRNAs need to be parametrized and more than 700 features have been described. Positive training examples for machine learning are readily available, but negative data is hard to come by. Therefore, it seems prerogative to use one-class classification instead of two-class classification. Previously, we were able to almost reach two-class classification accuracy using one-class classifiers. In this work, we employ feature selection procedures in conjunction with one-class classification and show that there is up to 36% difference in accuracy among these feature selection methods. The best feature set allowed the training of a one-class classifier which achieved an average accuracy of ~95.6% thereby outperforming previous two-class-based plant miRNA detection approaches by about 0.5%. We believe that this can be improved upon in the future by rigorous filtering of the positive training examples and by improving current feature clustering algorithms to better target pre-miRNA feature selection.

Word pair classification during imagined speech using direct brain recordings

NASA Astrophysics Data System (ADS)

Martin, Stephanie; Brunner, Peter; Iturrate, Iñaki; Millán, José Del R.; Schalk, Gerwin; Knight, Robert T.; Pasley, Brian N.

2016-05-01

People that cannot communicate due to neurological disorders would benefit from an internal speech decoder. Here, we showed the ability to classify individual words during imagined speech from electrocorticographic signals. In a word imagery task, we used high gamma (70-150 Hz) time features with a support vector machine model to classify individual words from a pair of words. To account for temporal irregularities during speech production, we introduced a non-linear time alignment into the SVM kernel. Classification accuracy reached 88% in a two-class classification framework (50% chance level), and average classification accuracy across fifteen word-pairs was significant across five subjects (mean = 58% p < 0.05). We also compared classification accuracy between imagined speech, overt speech and listening. As predicted, higher classification accuracy was obtained in the listening and overt speech conditions (mean = 89% and 86%, respectively; p < 0.0001), where speech stimuli were directly presented. The results provide evidence for a neural representation for imagined words in the temporal lobe, frontal lobe and sensorimotor cortex, consistent with previous findings in speech perception and production. These data represent a proof of concept study for basic decoding of speech imagery, and delineate a number of key challenges to usage of speech imagery neural representations for clinical applications.

Word pair classification during imagined speech using direct brain recordings

PubMed Central

Martin, Stephanie; Brunner, Peter; Iturrate, Iñaki; Millán, José del R.; Schalk, Gerwin; Knight, Robert T.; Pasley, Brian N.

2016-01-01

People that cannot communicate due to neurological disorders would benefit from an internal speech decoder. Here, we showed the ability to classify individual words during imagined speech from electrocorticographic signals. In a word imagery task, we used high gamma (70–150 Hz) time features with a support vector machine model to classify individual words from a pair of words. To account for temporal irregularities during speech production, we introduced a non-linear time alignment into the SVM kernel. Classification accuracy reached 88% in a two-class classification framework (50% chance level), and average classification accuracy across fifteen word-pairs was significant across five subjects (mean = 58%; p < 0.05). We also compared classification accuracy between imagined speech, overt speech and listening. As predicted, higher classification accuracy was obtained in the listening and overt speech conditions (mean = 89% and 86%, respectively; p < 0.0001), where speech stimuli were directly presented. The results provide evidence for a neural representation for imagined words in the temporal lobe, frontal lobe and sensorimotor cortex, consistent with previous findings in speech perception and production. These data represent a proof of concept study for basic decoding of speech imagery, and delineate a number of key challenges to usage of speech imagery neural representations for clinical applications. PMID:27165452

Comparing ecoregional classifications for natural areas management in the Klamath Region, USA

USGS Publications Warehouse

Sarr, Daniel A.; Duff, Andrew; Dinger, Eric C.; Shafer, Sarah L.; Wing, Michael; Seavy, Nathaniel E.; Alexander, John D.

2015-01-01

We compared three existing ecoregional classification schemes (Bailey, Omernik, and World Wildlife Fund) with two derived schemes (Omernik Revised and Climate Zones) to explore their effectiveness in explaining species distributions and to better understand natural resource geography in the Klamath Region, USA. We analyzed presence/absence data derived from digital distribution maps for trees, amphibians, large mammals, small mammals, migrant birds, and resident birds using three statistical analyses of classification accuracy (Analysis of Similarity, Canonical Analysis of Principal Coordinates, and Classification Strength). The classifications were roughly comparable in classification accuracy, with Omernik Revised showing the best overall performance. Trees showed the strongest fidelity to the classifications, and large mammals showed the weakest fidelity. We discuss the implications for regional biogeography and describe how intermediate resolution ecoregional classifications may be appropriate for use as natural areas management domains.

a Gsa-Svm Hybrid System for Classification of Binary Problems

NASA Astrophysics Data System (ADS)

Sarafrazi, Soroor; Nezamabadi-pour, Hossein; Barahman, Mojgan

2011-06-01

This paperhybridizesgravitational search algorithm (GSA) with support vector machine (SVM) and made a novel GSA-SVM hybrid system to improve the classification accuracy in binary problems. GSA is an optimization heuristic toolused to optimize the value of SVM kernel parameter (in this paper, radial basis function (RBF) is chosen as the kernel function). The experimental results show that this newapproach can achieve high classification accuracy and is comparable to or better than the particle swarm optimization (PSO)-SVM and genetic algorithm (GA)-SVM, which are two hybrid systems for classification.

Typicality effects in artificial categories: is there a hemisphere difference?

PubMed

Richards, L G; Chiarello, C

1990-07-01

In category classification tasks, typicality effects are usually found: accuracy and reaction time depend upon distance from a prototype. In this study, subjects learned either verbal or nonverbal dot pattern categories, followed by a lateralized classification task. Comparable typicality effects were found in both reaction time and accuracy across visual fields for both verbal and nonverbal categories. Both hemispheres appeared to use a similarity-to-prototype matching strategy in classification. This indicates that merely having a verbal label does not differentiate classification in the two hemispheres.

Multi-site evaluation of IKONOS data for classification of tropical coral reef environments

USGS Publications Warehouse

Andrefouet, S.; Kramer, Philip; Torres-Pulliza, D.; Joyce, K.E.; Hochberg, E.J.; Garza-Perez, R.; Mumby, P.J.; Riegl, Bernhard; Yamano, H.; White, W.H.; Zubia, M.; Brock, J.C.; Phinn, S.R.; Naseer, A.; Hatcher, B.G.; Muller-Karger, F. E.

2003-01-01

Ten IKONOS images of different coral reef sites distributed around the world were processed to assess the potential of 4-m resolution multispectral data for coral reef habitat mapping. Complexity of reef environments, established by field observation, ranged from 3 to 15 classes of benthic habitats containing various combinations of sediments, carbonate pavement, seagrass, algae, and corals in different geomorphologic zones (forereef, lagoon, patch reef, reef flats). Processing included corrections for sea surface roughness and bathymetry, unsupervised or supervised classification, and accuracy assessment based on ground-truth data. IKONOS classification results were compared with classified Landsat 7 imagery for simple to moderate complexity of reef habitats (5-11 classes). For both sensors, overall accuracies of the classifications show a general linear trend of decreasing accuracy with increasing habitat complexity. The IKONOS sensor performed better, with a 15-20% improvement in accuracy compared to Landsat. For IKONOS, overall accuracy was 77% for 4-5 classes, 71% for 7-8 classes, 65% in 9-11 classes, and 53% for more than 13 classes. The Landsat classification accuracy was systematically lower, with an average of 56% for 5-10 classes. Within this general trend, inter-site comparisons and specificities demonstrate the benefits of different approaches. Pre-segmentation of the different geomorphologic zones and depth correction provided different advantages in different environments. Our results help guide scientists and managers in applying IKONOS-class data for coral reef mapping applications. ?? 2003 Elsevier Inc. All rights reserved.

Transportation Modes Classification Using Sensors on Smartphones.

PubMed

Fang, Shih-Hau; Liao, Hao-Hsiang; Fei, Yu-Xiang; Chen, Kai-Hsiang; Huang, Jen-Wei; Lu, Yu-Ding; Tsao, Yu

2016-08-19

This paper investigates the transportation and vehicular modes classification by using big data from smartphone sensors. The three types of sensors used in this paper include the accelerometer, magnetometer, and gyroscope. This study proposes improved features and uses three machine learning algorithms including decision trees, K-nearest neighbor, and support vector machine to classify the user's transportation and vehicular modes. In the experiments, we discussed and compared the performance from different perspectives including the accuracy for both modes, the executive time, and the model size. Results show that the proposed features enhance the accuracy, in which the support vector machine provides the best performance in classification accuracy whereas it consumes the largest prediction time. This paper also investigates the vehicle classification mode and compares the results with that of the transportation modes.

Transportation Modes Classification Using Sensors on Smartphones

PubMed Central

Fang, Shih-Hau; Liao, Hao-Hsiang; Fei, Yu-Xiang; Chen, Kai-Hsiang; Huang, Jen-Wei; Lu, Yu-Ding; Tsao, Yu

2016-01-01

This paper investigates the transportation and vehicular modes classification by using big data from smartphone sensors. The three types of sensors used in this paper include the accelerometer, magnetometer, and gyroscope. This study proposes improved features and uses three machine learning algorithms including decision trees, K-nearest neighbor, and support vector machine to classify the user’s transportation and vehicular modes. In the experiments, we discussed and compared the performance from different perspectives including the accuracy for both modes, the executive time, and the model size. Results show that the proposed features enhance the accuracy, in which the support vector machine provides the best performance in classification accuracy whereas it consumes the largest prediction time. This paper also investigates the vehicle classification mode and compares the results with that of the transportation modes. PMID:27548182

Brain-Computer Interface Based on Generation of Visual Images

PubMed Central

Bobrov, Pavel; Frolov, Alexander; Cantor, Charles; Fedulova, Irina; Bakhnyan, Mikhail; Zhavoronkov, Alexander

2011-01-01

This paper examines the task of recognizing EEG patterns that correspond to performing three mental tasks: relaxation and imagining of two types of pictures: faces and houses. The experiments were performed using two EEG headsets: BrainProducts ActiCap and Emotiv EPOC. The Emotiv headset becomes widely used in consumer BCI application allowing for conducting large-scale EEG experiments in the future. Since classification accuracy significantly exceeded the level of random classification during the first three days of the experiment with EPOC headset, a control experiment was performed on the fourth day using ActiCap. The control experiment has shown that utilization of high-quality research equipment can enhance classification accuracy (up to 68% in some subjects) and that the accuracy is independent of the presence of EEG artifacts related to blinking and eye movement. This study also shows that computationally-inexpensive Bayesian classifier based on covariance matrix analysis yields similar classification accuracy in this problem as a more sophisticated Multi-class Common Spatial Patterns (MCSP) classifier. PMID:21695206

Comparison of MSS and TM Data for Landcover Classification in the Chesapeake Bay Area: a Preliminary Report. [Taylor's Island, Maryland

NASA Technical Reports Server (NTRS)

Mulligan, P. J.; Gervin, J. C.; Lu, Y. C.

1985-01-01

An area bordering the Eastern Shore of the Chesapeake Bay was selected for study and classified using unsupervised techniques applied to LANDSAT-2 MSS data and several band combinations of LANDSAT-4 TM data. The accuracies of these Level I land cover classifications were verified using the Taylor's Island USGS 7.5 minute topographic map which was photointerpreted, digitized and rasterized. The the Taylor's Island map, comparing the MSS and TM three band (2 3 4) classifications, the increased resolution of TM produced a small improvement in overall accuracy of 1% correct due primarily to a small improvement, and 1% and 3%, in areas such as water and woodland. This was expected as the MSS data typically produce high accuracies for categories which cover large contiguous areas. However, in the categories covering smaller areas within the map there was generally an improvement of at least 10%. Classification of the important residential category improved 12%, and wetlands were mapped with 11% greater accuracy.

Comparison Between Spectral, Spatial and Polarimetric Classification of Urban and Periurban Landcover Using Temporal Sentinel - 1 Images

NASA Astrophysics Data System (ADS)

Roychowdhury, K.

2016-06-01

Landcover is the easiest detectable indicator of human interventions on land. Urban and peri-urban areas present a complex combination of landcover, which makes classification challenging. This paper assesses the different methods of classifying landcover using dual polarimetric Sentinel-1 data collected during monsoon (July) and winter (December) months of 2015. Four broad landcover classes such as built up areas, water bodies and wetlands, vegetation and open spaces of Kolkata and its surrounding regions were identified. Polarimetric analyses were conducted on Single Look Complex (SLC) data of the region while ground range detected (GRD) data were used for spectral and spatial classification. Unsupervised classification by means of K-Means clustering used backscatter values and was able to identify homogenous landcovers over the study area. The results produced an overall accuracy of less than 50% for both the seasons. Higher classification accuracy (around 70%) was achieved by adding texture variables as inputs along with the backscatter values. However, the accuracy of classification increased significantly with polarimetric analyses. The overall accuracy was around 80% in Wishart H-A-Alpha unsupervised classification. The method was useful in identifying urban areas due to their double-bounce scattering and vegetated areas, which have more random scattering. Normalized Difference Built-up index (NDBI) and Normalized Difference Vegetation Index (NDVI) obtained from Landsat 8 data over the study area were used to verify vegetation and urban classes. The study compares the accuracies of different methods of classifying landcover using medium resolution SAR data in a complex urban area and suggests that polarimetric analyses present the most accurate results for urban and suburban areas.

Detection of colorectal masses in CT colonography: application of deep residual networks for differentiating masses from normal colon anatomy

NASA Astrophysics Data System (ADS)

Näppi, Janne J.; Hironaka, Toru; Yoshida, Hiroyuki

2018-02-01

Even though the clinical consequences of a missed colorectal cancer far outweigh those of a missed polyp, there has been little work on computer-aided detection (CADe) for colorectal masses in CT colonography (CTC). One of the problems is that it is not clear how to manually design mathematical image-based features that could be used to differentiate effectively between masses and certain types of normal colon anatomy such as ileocecal valves (ICVs). Deep learning has demonstrated ability to automatically determine effective discriminating features in many image-based problems. Recently, residual networks (ResNets) were developed to address the practical problems of constructing deep network architectures for optimizing the performance of deep learning. In this pilot study, we compared the classification performance of a conventional 2D-convolutional ResNet (2D-ResNet) with that of a volumetric 3D-convolutional ResNet (3D-ResNet) in differentiating masses from normal colon anatomy in CTC. For the development and evaluation of the ResNets, 695 volumetric images of biopsy-proven colorectal masses, ICVs, haustral folds, and rectal tubes were sampled from 196 clinical CTC cases and divided randomly into independent training, validation, and test datasets. The training set was expanded by use of volumetric data augmentation. Our preliminary results on the 140 test samples indicate that it is feasible to train a deep volumetric 3D-ResNet for performing effective image-based discriminations in CTC. The 3D-ResNet slightly outperformed the 2D-ResNet in the discrimination of masses and normal colon anatomy, but the statistical difference between their very high classification accuracies was not significant. The highest classification accuracy was obtained by combining the mass-likelihood estimates of the 2D- and 3D-ResNets, which enabled correct classification of all of the masses.

Brain Gliomas: Multicenter Standardized Assessment of Dynamic Contrast-enhanced and Dynamic Susceptibility Contrast MR Images.

PubMed

Anzalone, Nicoletta; Castellano, Antonella; Cadioli, Marcello; Conte, Gian Marco; Cuccarini, Valeria; Bizzi, Alberto; Grimaldi, Marco; Costa, Antonella; Grillea, Giovanni; Vitali, Paolo; Aquino, Domenico; Terreni, Maria Rosa; Torri, Valter; Erickson, Bradley J; Caulo, Massimo

2018-06-01

Purpose To evaluate the feasibility of a standardized protocol for acquisition and analysis of dynamic contrast material-enhanced (DCE) and dynamic susceptibility contrast (DSC) magnetic resonance (MR) imaging in a multicenter clinical setting and to verify its accuracy in predicting glioma grade according to the new World Health Organization 2016 classification. Materials and Methods The local research ethics committees of all centers approved the study, and informed consent was obtained from patients. One hundred patients with glioma were prospectively examined at 3.0 T in seven centers that performed the same preoperative MR imaging protocol, including DCE and DSC sequences. Two independent readers identified the perfusion hotspots on maps of volume transfer constant (K trans ), plasma (v p ) and extravascular-extracellular space (v e ) volumes, initial area under the concentration curve, and relative cerebral blood volume (rCBV). Differences in parameters between grades and molecular subtypes were assessed by using Kruskal-Wallis and Mann-Whitney U tests. Diagnostic accuracy was evaluated by using receiver operating characteristic curve analysis. Results The whole protocol was tolerated in all patients. Perfusion maps were successfully obtained in 94 patients. An excellent interreader reproducibility of DSC- and DCE-derived measures was found. Among DCE-derived parameters, v p and v e had the highest accuracy (are under the receiver operating characteristic curve [A z ] = 0.847 and 0.853) for glioma grading. DSC-derived rCBV had the highest accuracy (A z = 0.894), but the difference was not statistically significant (P > .05). Among lower-grade gliomas, a moderate increase in both v p and rCBV was evident in isocitrate dehydrogenase wild-type tumors, although this was not significant (P > .05). Conclusion A standardized multicenter acquisition and analysis protocol of DCE and DSC MR imaging is feasible and highly reproducible. Both techniques showed a comparable, high diagnostic accuracy for grading gliomas. © RSNA, 2018 Online supplemental material is available for this article.

An empirical comparison of different approaches for combining multimodal neuroimaging data with support vector machine

PubMed Central

Pettersson-Yeo, William; Benetti, Stefania; Marquand, Andre F.; Joules, Richard; Catani, Marco; Williams, Steve C. R.; Allen, Paul; McGuire, Philip; Mechelli, Andrea

2014-01-01

In the pursuit of clinical utility, neuroimaging researchers of psychiatric and neurological illness are increasingly using analyses, such as support vector machine, that allow inference at the single-subject level. Recent studies employing single-modality data, however, suggest that classification accuracies must be improved for such utility to be realized. One possible solution is to integrate different data types to provide a single combined output classification; either by generating a single decision function based on an integrated kernel matrix, or, by creating an ensemble of multiple single modality classifiers and integrating their predictions. Here, we describe four integrative approaches: (1) an un-weighted sum of kernels, (2) multi-kernel learning, (3) prediction averaging, and (4) majority voting, and compare their ability to enhance classification accuracy relative to the best single-modality classification accuracy. We achieve this by integrating structural, functional, and diffusion tensor magnetic resonance imaging data, in order to compare ultra-high risk (n = 19), first episode psychosis (n = 19) and healthy control subjects (n = 23). Our results show that (i) whilst integration can enhance classification accuracy by up to 13%, the frequency of such instances may be limited, (ii) where classification can be enhanced, simple methods may yield greater increases relative to more computationally complex alternatives, and, (iii) the potential for classification enhancement is highly influenced by the specific diagnostic comparison under consideration. In conclusion, our findings suggest that for moderately sized clinical neuroimaging datasets, combining different imaging modalities in a data-driven manner is no “magic bullet” for increasing classification accuracy. However, it remains possible that this conclusion is dependent on the use of neuroimaging modalities that had little, or no, complementary information to offer one another, and that the integration of more diverse types of data would have produced greater classification enhancement. We suggest that future studies ideally examine a greater variety of data types (e.g., genetic, cognitive, and neuroimaging) in order to identify the data types and combinations optimally suited to the classification of early stage psychosis. PMID:25076868

An empirical comparison of different approaches for combining multimodal neuroimaging data with support vector machine.

PubMed

Pettersson-Yeo, William; Benetti, Stefania; Marquand, Andre F; Joules, Richard; Catani, Marco; Williams, Steve C R; Allen, Paul; McGuire, Philip; Mechelli, Andrea

2014-01-01

In the pursuit of clinical utility, neuroimaging researchers of psychiatric and neurological illness are increasingly using analyses, such as support vector machine, that allow inference at the single-subject level. Recent studies employing single-modality data, however, suggest that classification accuracies must be improved for such utility to be realized. One possible solution is to integrate different data types to provide a single combined output classification; either by generating a single decision function based on an integrated kernel matrix, or, by creating an ensemble of multiple single modality classifiers and integrating their predictions. Here, we describe four integrative approaches: (1) an un-weighted sum of kernels, (2) multi-kernel learning, (3) prediction averaging, and (4) majority voting, and compare their ability to enhance classification accuracy relative to the best single-modality classification accuracy. We achieve this by integrating structural, functional, and diffusion tensor magnetic resonance imaging data, in order to compare ultra-high risk (n = 19), first episode psychosis (n = 19) and healthy control subjects (n = 23). Our results show that (i) whilst integration can enhance classification accuracy by up to 13%, the frequency of such instances may be limited, (ii) where classification can be enhanced, simple methods may yield greater increases relative to more computationally complex alternatives, and, (iii) the potential for classification enhancement is highly influenced by the specific diagnostic comparison under consideration. In conclusion, our findings suggest that for moderately sized clinical neuroimaging datasets, combining different imaging modalities in a data-driven manner is no "magic bullet" for increasing classification accuracy. However, it remains possible that this conclusion is dependent on the use of neuroimaging modalities that had little, or no, complementary information to offer one another, and that the integration of more diverse types of data would have produced greater classification enhancement. We suggest that future studies ideally examine a greater variety of data types (e.g., genetic, cognitive, and neuroimaging) in order to identify the data types and combinations optimally suited to the classification of early stage psychosis.

Mapping Species Composition of Forests and Tree Plantations in Northeastern Costa Rica with an Integration of Hyperspectral and Multitemporal Landsat Imagery

NASA Technical Reports Server (NTRS)

Fagan, Matthew E.; Defries, Ruth S.; Sesnie, Steven E.; Arroyo-Mora, J. Pablo; Soto, Carlomagno; Singh, Aditya; Townsend, Philip A.; Chazdon, Robin L.

2015-01-01

An efficient means to map tree plantations is needed to detect tropical land use change and evaluate reforestation projects. To analyze recent tree plantation expansion in northeastern Costa Rica, we examined the potential of combining moderate-resolution hyperspectral imagery (2005 HyMap mosaic) with multitemporal, multispectral data (Landsat) to accurately classify (1) general forest types and (2) tree plantations by species composition. Following a linear discriminant analysis to reduce data dimensionality, we compared four Random Forest classification models: hyperspectral data (HD) alone; HD plus interannual spectral metrics; HD plus a multitemporal forest regrowth classification; and all three models combined. The fourth, combined model achieved overall accuracy of 88.5%. Adding multitemporal data significantly improved classification accuracy (p less than 0.0001) of all forest types, although the effect on tree plantation accuracy was modest. The hyperspectral data alone classified six species of tree plantations with 75% to 93% producer's accuracy; adding multitemporal spectral data increased accuracy only for two species with dense canopies. Non-native tree species had higher classification accuracy overall and made up the majority of tree plantations in this landscape. Our results indicate that combining occasionally acquired hyperspectral data with widely available multitemporal satellite imagery enhances mapping and monitoring of reforestation in tropical landscapes.

Study of USGS/NASA land use classification system. [compatibility of land use classification system with computer processing techniques employed for land use mapping from ERTS data

NASA Technical Reports Server (NTRS)

Spann, G. W.; Faust, N. L.

1974-01-01

It is known from several previous investigations that many categories of land-use can be mapped via computer processing of Earth Resources Technology Satellite data. The results are presented of one such experiment using the USGS/NASA land-use classification system. Douglas County, Georgia, was chosen as the test site for this project. It was chosen primarily because of its recent rapid growth and future growth potential. Results of the investigation indicate an overall land-use mapping accuracy of 67% with higher accuracies in rural areas and lower accuracies in urban areas. It is estimated, however, that 95% of the State of Georgia could be mapped by these techniques with an accuracy of 80% to 90%.

Bayesian Network Structure Learning for Urban Land Use Classification from Landsat ETM+ and Ancillary Data

NASA Astrophysics Data System (ADS)

Park, M.; Stenstrom, M. K.

2004-12-01

Recognizing urban information from the satellite imagery is problematic due to the diverse features and dynamic changes of urban landuse. The use of Landsat imagery for urban land use classification involves inherent uncertainty due to its spatial resolution and the low separability among land uses. To resolve the uncertainty problem, we investigated the performance of Bayesian networks to classify urban land use since Bayesian networks provide a quantitative way of handling uncertainty and have been successfully used in many areas. In this study, we developed the optimized networks for urban land use classification from Landsat ETM+ images of Marina del Rey area based on USGS land cover/use classification level III. The networks started from a tree structure based on mutual information between variables and added the links to improve accuracy. This methodology offers several advantages: (1) The network structure shows the dependency relationships between variables. The class node value can be predicted even with particular band information missing due to sensor system error. The missing information can be inferred from other dependent bands. (2) The network structure provides information of variables that are important for the classification, which is not available from conventional classification methods such as neural networks and maximum likelihood classification. In our case, for example, bands 1, 5 and 6 are the most important inputs in determining the land use of each pixel. (3) The networks can be reduced with those input variables important for classification. This minimizes the problem without considering all possible variables. We also examined the effect of incorporating ancillary data: geospatial information such as X and Y coordinate values of each pixel and DEM data, and vegetation indices such as NDVI and Tasseled Cap transformation. The results showed that the locational information improved overall accuracy (81%) and kappa coefficient (76%), and lowered the omission and commission errors compared with using only spectral data (accuracy 71%, kappa coefficient 62%). Incorporating DEM data did not significantly improve overall accuracy (74%) and kappa coefficient (66%) but lowered the omission and commission errors. Incorporating NDVI did not much improve the overall accuracy (72%) and k coefficient (65%). Including Tasseled Cap transformation reduced the accuracy (accuracy 70%, kappa 61%). Therefore, additional information from the DEM and vegetation indices was not useful as locational ancillary data.

Steady-state visual evoked potential (SSVEP)-based communication: impact of harmonic frequency components

NASA Astrophysics Data System (ADS)

Müller-Putz, Gernot R.; Scherer, Reinhold; Brauneis, Christian; Pfurtscheller, Gert

2005-12-01

Brain-computer interfaces (BCIs) can be realized on the basis of steady-state evoked potentials (SSEPs). These types of brain signals resulting from repetitive stimulation have the same fundamental frequency as the stimulation but also include higher harmonics. This study investigated how the classification accuracy of a 4-class BCI system can be improved by incorporating visually evoked harmonic oscillations. The current study revealed that the use of three SSVEP harmonics yielded a significantly higher classification accuracy than was the case for one or two harmonics. During feedback experiments, the five subjects investigated reached a classification accuracy between 42.5% and 94.4%.

Steady-state visual evoked potential (SSVEP)-based communication: impact of harmonic frequency components.

PubMed

Müller-Putz, Gernot R; Scherer, Reinhold; Brauneis, Christian; Pfurtscheller, Gert

2005-12-01

Brain-computer interfaces (BCIs) can be realized on the basis of steady-state evoked potentials (SSEPs). These types of brain signals resulting from repetitive stimulation have the same fundamental frequency as the stimulation but also include higher harmonics. This study investigated how the classification accuracy of a 4-class BCI system can be improved by incorporating visually evoked harmonic oscillations. The current study revealed that the use of three SSVEP harmonics yielded a significantly higher classification accuracy than was the case for one or two harmonics. During feedback experiments, the five subjects investigated reached a classification accuracy between 42.5% and 94.4%.

Dynamic and scalable audio classification by collective network of binary classifiers framework: an evolutionary approach.

PubMed

Kiranyaz, Serkan; Mäkinen, Toni; Gabbouj, Moncef

2012-10-01

In this paper, we propose a novel framework based on a collective network of evolutionary binary classifiers (CNBC) to address the problems of feature and class scalability. The main goal of the proposed framework is to achieve a high classification performance over dynamic audio and video repositories. The proposed framework adopts a "Divide and Conquer" approach in which an individual network of binary classifiers (NBC) is allocated to discriminate each audio class. An evolutionary search is applied to find the best binary classifier in each NBC with respect to a given criterion. Through the incremental evolution sessions, the CNBC framework can dynamically adapt to each new incoming class or feature set without resorting to a full-scale re-training or re-configuration. Therefore, the CNBC framework is particularly designed for dynamically varying databases where no conventional static classifiers can adapt to such changes. In short, it is entirely a novel topology, an unprecedented approach for dynamic, content/data adaptive and scalable audio classification. A large set of audio features can be effectively used in the framework, where the CNBCs make appropriate selections and combinations so as to achieve the highest discrimination among individual audio classes. Experiments demonstrate a high classification accuracy (above 90%) and efficiency of the proposed framework over large and dynamic audio databases. Copyright © 2012 Elsevier Ltd. All rights reserved.

CFS-SMO based classification of breast density using multiple texture models.

PubMed

Sharma, Vipul; Singh, Sukhwinder

2014-06-01

It is highly acknowledged in the medical profession that density of breast tissue is a major cause for the growth of breast cancer. Increased breast density was found to be linked with an increased risk of breast cancer growth, as high density makes it difficult for radiologists to see an abnormality which leads to false negative results. Therefore, there is need for the development of highly efficient techniques for breast tissue classification based on density. This paper presents a hybrid scheme for classification of fatty and dense mammograms using correlation-based feature selection (CFS) and sequential minimal optimization (SMO). In this work, texture analysis is done on a region of interest selected from the mammogram. Various texture models have been used to quantify the texture of parenchymal patterns of breast. To reduce the dimensionality and to identify the features which differentiate between breast tissue densities, CFS is used. Finally, classification is performed using SMO. The performance is evaluated using 322 images of mini-MIAS database. Highest accuracy of 96.46% is obtained for two-class problem (fatty and dense) using proposed approach. Performance of selected features by CFS is also evaluated by Naïve Bayes, Multilayer Perceptron, RBF Network, J48 and kNN classifier. The proposed CFS-SMO method outperforms all other classifiers giving a sensitivity of 100%. This makes it suitable to be taken as a second opinion in classifying breast tissue density.

Classifying depression patients and normal subjects using machine learning techniques and nonlinear features from EEG signal.

PubMed

Hosseinifard, Behshad; Moradi, Mohammad Hassan; Rostami, Reza

2013-03-01

Diagnosing depression in the early curable stages is very important and may even save the life of a patient. In this paper, we study nonlinear analysis of EEG signal for discriminating depression patients and normal controls. Forty-five unmedicated depressed patients and 45 normal subjects were participated in this study. Power of four EEG bands and four nonlinear features including detrended fluctuation analysis (DFA), higuchi fractal, correlation dimension and lyapunov exponent were extracted from EEG signal. For discriminating the two groups, k-nearest neighbor, linear discriminant analysis and logistic regression as the classifiers are then used. Highest classification accuracy of 83.3% is obtained by correlation dimension and LR classifier among other nonlinear features. For further improvement, all nonlinear features are combined and applied to classifiers. A classification accuracy of 90% is achieved by all nonlinear features and LR classifier. In all experiments, genetic algorithm is employed to select the most important features. The proposed technique is compared and contrasted with the other reported methods and it is demonstrated that by combining nonlinear features, the performance is enhanced. This study shows that nonlinear analysis of EEG can be a useful method for discriminating depressed patients and normal subjects. It is suggested that this analysis may be a complementary tool to help psychiatrists for diagnosing depressed patients. Copyright © 2012 Elsevier Ireland Ltd. All rights reserved.

High Accuracy Human Activity Recognition Based on Sparse Locality Preserving Projections.

PubMed

Zhu, Xiangbin; Qiu, Huiling

2016-01-01

Human activity recognition(HAR) from the temporal streams of sensory data has been applied to many fields, such as healthcare services, intelligent environments and cyber security. However, the classification accuracy of most existed methods is not enough in some applications, especially for healthcare services. In order to improving accuracy, it is necessary to develop a novel method which will take full account of the intrinsic sequential characteristics for time-series sensory data. Moreover, each human activity may has correlated feature relationship at different levels. Therefore, in this paper, we propose a three-stage continuous hidden Markov model (TSCHMM) approach to recognize human activities. The proposed method contains coarse, fine and accurate classification. The feature reduction is an important step in classification processing. In this paper, sparse locality preserving projections (SpLPP) is exploited to determine the optimal feature subsets for accurate classification of the stationary-activity data. It can extract more discriminative activities features from the sensor data compared with locality preserving projections. Furthermore, all of the gyro-based features are used for accurate classification of the moving data. Compared with other methods, our method uses significantly less number of features, and the over-all accuracy has been obviously improved.

High Accuracy Human Activity Recognition Based on Sparse Locality Preserving Projections

PubMed Central

2016-01-01

Human activity recognition(HAR) from the temporal streams of sensory data has been applied to many fields, such as healthcare services, intelligent environments and cyber security. However, the classification accuracy of most existed methods is not enough in some applications, especially for healthcare services. In order to improving accuracy, it is necessary to develop a novel method which will take full account of the intrinsic sequential characteristics for time-series sensory data. Moreover, each human activity may has correlated feature relationship at different levels. Therefore, in this paper, we propose a three-stage continuous hidden Markov model (TSCHMM) approach to recognize human activities. The proposed method contains coarse, fine and accurate classification. The feature reduction is an important step in classification processing. In this paper, sparse locality preserving projections (SpLPP) is exploited to determine the optimal feature subsets for accurate classification of the stationary-activity data. It can extract more discriminative activities features from the sensor data compared with locality preserving projections. Furthermore, all of the gyro-based features are used for accurate classification of the moving data. Compared with other methods, our method uses significantly less number of features, and the over-all accuracy has been obviously improved. PMID:27893761

Additional studies of forest classification accuracy as influenced by multispectral scanner spatial resolution

NASA Technical Reports Server (NTRS)

Sadowski, F. E.; Sarno, J. E.

1976-01-01

First, an analysis of forest feature signatures was used to help explain the large variation in classification accuracy that can occur among individual forest features for any one case of spatial resolution and the inconsistent changes in classification accuracy that were demonstrated among features as spatial resolution was degraded. Second, the classification rejection threshold was varied in an effort to reduce the large proportion of unclassified resolution elements that previously appeared in the processing of coarse resolution data when a constant rejection threshold was used for all cases of spatial resolution. For the signature analysis, two-channel ellipse plots showing the feature signature distributions for several cases of spatial resolution indicated that the capability of signatures to correctly identify their respective features is dependent on the amount of statistical overlap among signatures. Reductions in signature variance that occur in data of degraded spatial resolution may not necessarily decrease the amount of statistical overlap among signatures having large variance and small mean separations. Features classified by such signatures may thus continue to have similar amounts of misclassified elements in coarser resolution data, and thus, not necessarily improve in classification accuracy.

Training sample selection based on self-training for liver cirrhosis classification using ultrasound images

NASA Astrophysics Data System (ADS)

Fujita, Yusuke; Mitani, Yoshihiro; Hamamoto, Yoshihiko; Segawa, Makoto; Terai, Shuji; Sakaida, Isao

2017-03-01

Ultrasound imaging is a popular and non-invasive tool used in the diagnoses of liver disease. Cirrhosis is a chronic liver disease and it can advance to liver cancer. Early detection and appropriate treatment are crucial to prevent liver cancer. However, ultrasound image analysis is very challenging, because of the low signal-to-noise ratio of ultrasound images. To achieve the higher classification performance, selection of training regions of interest (ROIs) is very important that effect to classification accuracy. The purpose of our study is cirrhosis detection with high accuracy using liver ultrasound images. In our previous works, training ROI selection by MILBoost and multiple-ROI classification based on the product rule had been proposed, to achieve high classification performance. In this article, we propose self-training method to select training ROIs effectively. Evaluation experiments were performed to evaluate effect of self-training, using manually selected ROIs and also automatically selected ROIs. Experimental results show that self-training for manually selected ROIs achieved higher classification performance than other approaches, including our conventional methods. The manually ROI definition and sample selection are important to improve classification accuracy in cirrhosis detection using ultrasound images.

The impact of OCR accuracy on automated cancer classification of pathology reports.

PubMed

Zuccon, Guido; Nguyen, Anthony N; Bergheim, Anton; Wickman, Sandra; Grayson, Narelle

2012-01-01

To evaluate the effects of Optical Character Recognition (OCR) on the automatic cancer classification of pathology reports. Scanned images of pathology reports were converted to electronic free-text using a commercial OCR system. A state-of-the-art cancer classification system, the Medical Text Extraction (MEDTEX) system, was used to automatically classify the OCR reports. Classifications produced by MEDTEX on the OCR versions of the reports were compared with the classification from a human amended version of the OCR reports. The employed OCR system was found to recognise scanned pathology reports with up to 99.12% character accuracy and up to 98.95% word accuracy. Errors in the OCR processing were found to minimally impact on the automatic classification of scanned pathology reports into notifiable groups. However, the impact of OCR errors is not negligible when considering the extraction of cancer notification items, such as primary site, histological type, etc. The automatic cancer classification system used in this work, MEDTEX, has proven to be robust to errors produced by the acquisition of freetext pathology reports from scanned images through OCR software. However, issues emerge when considering the extraction of cancer notification items.

An experimental study of interstitial lung tissue classification in HRCT images using ANN and role of cost functions

NASA Astrophysics Data System (ADS)

Dash, Jatindra K.; Kale, Mandar; Mukhopadhyay, Sudipta; Khandelwal, Niranjan; Prabhakar, Nidhi; Garg, Mandeep; Kalra, Naveen

2017-03-01

In this paper, we investigate the effect of the error criteria used during a training phase of the artificial neural network (ANN) on the accuracy of the classifier for classification of lung tissues affected with Interstitial Lung Diseases (ILD). Mean square error (MSE) and the cross-entropy (CE) criteria are chosen being most popular choice in state-of-the-art implementations. The classification experiment performed on the six interstitial lung disease (ILD) patterns viz. Consolidation, Emphysema, Ground Glass Opacity, Micronodules, Fibrosis and Healthy from MedGIFT database. The texture features from an arbitrary region of interest (AROI) are extracted using Gabor filter. Two different neural networks are trained with the scaled conjugate gradient back propagation algorithm with MSE and CE error criteria function respectively for weight updation. Performance is evaluated in terms of average accuracy of these classifiers using 4 fold cross-validation. Each network is trained for five times for each fold with randomly initialized weight vectors and accuracies are computed. Significant improvement in classification accuracy is observed when ANN is trained by using CE (67.27%) as error function compared to MSE (63.60%). Moreover, standard deviation of the classification accuracy for the network trained with CE (6.69) error criteria is found less as compared to network trained with MSE (10.32) criteria.

Big Data: A Parallel Particle Swarm Optimization-Back-Propagation Neural Network Algorithm Based on MapReduce.

PubMed

Cao, Jianfang; Cui, Hongyan; Shi, Hao; Jiao, Lijuan

2016-01-01

A back-propagation (BP) neural network can solve complicated random nonlinear mapping problems; therefore, it can be applied to a wide range of problems. However, as the sample size increases, the time required to train BP neural networks becomes lengthy. Moreover, the classification accuracy decreases as well. To improve the classification accuracy and runtime efficiency of the BP neural network algorithm, we proposed a parallel design and realization method for a particle swarm optimization (PSO)-optimized BP neural network based on MapReduce on the Hadoop platform using both the PSO algorithm and a parallel design. The PSO algorithm was used to optimize the BP neural network's initial weights and thresholds and improve the accuracy of the classification algorithm. The MapReduce parallel programming model was utilized to achieve parallel processing of the BP algorithm, thereby solving the problems of hardware and communication overhead when the BP neural network addresses big data. Datasets on 5 different scales were constructed using the scene image library from the SUN Database. The classification accuracy of the parallel PSO-BP neural network algorithm is approximately 92%, and the system efficiency is approximately 0.85, which presents obvious advantages when processing big data. The algorithm proposed in this study demonstrated both higher classification accuracy and improved time efficiency, which represents a significant improvement obtained from applying parallel processing to an intelligent algorithm on big data.

Effects of sample survey design on the accuracy of classification tree models in species distribution models

USGS Publications Warehouse

Edwards, T.C.; Cutler, D.R.; Zimmermann, N.E.; Geiser, L.; Moisen, Gretchen G.

2006-01-01

We evaluated the effects of probabilistic (hereafter DESIGN) and non-probabilistic (PURPOSIVE) sample surveys on resultant classification tree models for predicting the presence of four lichen species in the Pacific Northwest, USA. Models derived from both survey forms were assessed using an independent data set (EVALUATION). Measures of accuracy as gauged by resubstitution rates were similar for each lichen species irrespective of the underlying sample survey form. Cross-validation estimates of prediction accuracies were lower than resubstitution accuracies for all species and both design types, and in all cases were closer to the true prediction accuracies based on the EVALUATION data set. We argue that greater emphasis should be placed on calculating and reporting cross-validation accuracy rates rather than simple resubstitution accuracy rates. Evaluation of the DESIGN and PURPOSIVE tree models on the EVALUATION data set shows significantly lower prediction accuracy for the PURPOSIVE tree models relative to the DESIGN models, indicating that non-probabilistic sample surveys may generate models with limited predictive capability. These differences were consistent across all four lichen species, with 11 of the 12 possible species and sample survey type comparisons having significantly lower accuracy rates. Some differences in accuracy were as large as 50%. The classification tree structures also differed considerably both among and within the modelled species, depending on the sample survey form. Overlap in the predictor variables selected by the DESIGN and PURPOSIVE tree models ranged from only 20% to 38%, indicating the classification trees fit the two evaluated survey forms on different sets of predictor variables. The magnitude of these differences in predictor variables throws doubt on ecological interpretation derived from prediction models based on non-probabilistic sample surveys. ?? 2006 Elsevier B.V. All rights reserved.

Object-based land cover classification based on fusion of multifrequency SAR data and THAICHOTE optical imagery

NASA Astrophysics Data System (ADS)

Sukawattanavijit, Chanika; Srestasathiern, Panu

2017-10-01

Land Use and Land Cover (LULC) information are significant to observe and evaluate environmental change. LULC classification applying remotely sensed data is a technique popularly employed on a global and local dimension particularly, in urban areas which have diverse land cover types. These are essential components of the urban terrain and ecosystem. In the present, object-based image analysis (OBIA) is becoming widely popular for land cover classification using the high-resolution image. COSMO-SkyMed SAR data was fused with THAICHOTE (namely, THEOS: Thailand Earth Observation Satellite) optical data for land cover classification using object-based. This paper indicates a comparison between object-based and pixel-based approaches in image fusion. The per-pixel method, support vector machines (SVM) was implemented to the fused image based on Principal Component Analysis (PCA). For the objectbased classification was applied to the fused images to separate land cover classes by using nearest neighbor (NN) classifier. Finally, the accuracy assessment was employed by comparing with the classification of land cover mapping generated from fused image dataset and THAICHOTE image. The object-based data fused COSMO-SkyMed with THAICHOTE images demonstrated the best classification accuracies, well over 85%. As the results, an object-based data fusion provides higher land cover classification accuracy than per-pixel data fusion.

Comparative Study of SVM Methods Combined with Voxel Selection for Object Category Classification on fMRI Data

PubMed Central

Song, Sutao; Zhan, Zhichao; Long, Zhiying; Zhang, Jiacai; Yao, Li

2011-01-01

Background Support vector machine (SVM) has been widely used as accurate and reliable method to decipher brain patterns from functional MRI (fMRI) data. Previous studies have not found a clear benefit for non-linear (polynomial kernel) SVM versus linear one. Here, a more effective non-linear SVM using radial basis function (RBF) kernel is compared with linear SVM. Different from traditional studies which focused either merely on the evaluation of different types of SVM or the voxel selection methods, we aimed to investigate the overall performance of linear and RBF SVM for fMRI classification together with voxel selection schemes on classification accuracy and time-consuming. Methodology/Principal Findings Six different voxel selection methods were employed to decide which voxels of fMRI data would be included in SVM classifiers with linear and RBF kernels in classifying 4-category objects. Then the overall performances of voxel selection and classification methods were compared. Results showed that: (1) Voxel selection had an important impact on the classification accuracy of the classifiers: in a relative low dimensional feature space, RBF SVM outperformed linear SVM significantly; in a relative high dimensional space, linear SVM performed better than its counterpart; (2) Considering the classification accuracy and time-consuming holistically, linear SVM with relative more voxels as features and RBF SVM with small set of voxels (after PCA) could achieve the better accuracy and cost shorter time. Conclusions/Significance The present work provides the first empirical result of linear and RBF SVM in classification of fMRI data, combined with voxel selection methods. Based on the findings, if only classification accuracy was concerned, RBF SVM with appropriate small voxels and linear SVM with relative more voxels were two suggested solutions; if users concerned more about the computational time, RBF SVM with relative small set of voxels when part of the principal components were kept as features was a better choice. PMID:21359184

Comparative study of SVM methods combined with voxel selection for object category classification on fMRI data.

PubMed

Song, Sutao; Zhan, Zhichao; Long, Zhiying; Zhang, Jiacai; Yao, Li

2011-02-16

Support vector machine (SVM) has been widely used as accurate and reliable method to decipher brain patterns from functional MRI (fMRI) data. Previous studies have not found a clear benefit for non-linear (polynomial kernel) SVM versus linear one. Here, a more effective non-linear SVM using radial basis function (RBF) kernel is compared with linear SVM. Different from traditional studies which focused either merely on the evaluation of different types of SVM or the voxel selection methods, we aimed to investigate the overall performance of linear and RBF SVM for fMRI classification together with voxel selection schemes on classification accuracy and time-consuming. Six different voxel selection methods were employed to decide which voxels of fMRI data would be included in SVM classifiers with linear and RBF kernels in classifying 4-category objects. Then the overall performances of voxel selection and classification methods were compared. Results showed that: (1) Voxel selection had an important impact on the classification accuracy of the classifiers: in a relative low dimensional feature space, RBF SVM outperformed linear SVM significantly; in a relative high dimensional space, linear SVM performed better than its counterpart; (2) Considering the classification accuracy and time-consuming holistically, linear SVM with relative more voxels as features and RBF SVM with small set of voxels (after PCA) could achieve the better accuracy and cost shorter time. The present work provides the first empirical result of linear and RBF SVM in classification of fMRI data, combined with voxel selection methods. Based on the findings, if only classification accuracy was concerned, RBF SVM with appropriate small voxels and linear SVM with relative more voxels were two suggested solutions; if users concerned more about the computational time, RBF SVM with relative small set of voxels when part of the principal components were kept as features was a better choice.

Object-based random forest classification of Landsat ETM+ and WorldView-2 satellite imagery for mapping lowland native grassland communities in Tasmania, Australia

NASA Astrophysics Data System (ADS)

Melville, Bethany; Lucieer, Arko; Aryal, Jagannath

2018-04-01

This paper presents a random forest classification approach for identifying and mapping three types of lowland native grassland communities found in the Tasmanian Midlands region. Due to the high conservation priority assigned to these communities, there has been an increasing need to identify appropriate datasets that can be used to derive accurate and frequently updateable maps of community extent. Therefore, this paper proposes a method employing repeat classification and statistical significance testing as a means of identifying the most appropriate dataset for mapping these communities. Two datasets were acquired and analysed; a Landsat ETM+ scene, and a WorldView-2 scene, both from 2010. Training and validation data were randomly subset using a k-fold (k = 50) approach from a pre-existing field dataset. Poa labillardierei, Themeda triandra and lowland native grassland complex communities were identified in addition to dry woodland and agriculture. For each subset of randomly allocated points, a random forest model was trained based on each dataset, and then used to classify the corresponding imagery. Validation was performed using the reciprocal points from the independent subset that had not been used to train the model. Final training and classification accuracies were reported as per class means for each satellite dataset. Analysis of Variance (ANOVA) was undertaken to determine whether classification accuracy differed between the two datasets, as well as between classifications. Results showed mean class accuracies between 54% and 87%. Class accuracy only differed significantly between datasets for the dry woodland and Themeda grassland classes, with the WorldView-2 dataset showing higher mean classification accuracies. The results of this study indicate that remote sensing is a viable method for the identification of lowland native grassland communities in the Tasmanian Midlands, and that repeat classification and statistical significant testing can be used to identify optimal datasets for vegetation community mapping.

A wavelet-based technique to predict treatment outcome for Major Depressive Disorder

PubMed Central

Xia, Likun; Mohd Yasin, Mohd Azhar; Azhar Ali, Syed Saad

2017-01-01

Treatment management for Major Depressive Disorder (MDD) has been challenging. However, electroencephalogram (EEG)-based predictions of antidepressant’s treatment outcome may help during antidepressant’s selection and ultimately improve the quality of life for MDD patients. In this study, a machine learning (ML) method involving pretreatment EEG data was proposed to perform such predictions for Selective Serotonin Reuptake Inhibitor (SSRIs). For this purpose, the acquisition of experimental data involved 34 MDD patients and 30 healthy controls. Consequently, a feature matrix was constructed involving time-frequency decomposition of EEG data based on wavelet transform (WT) analysis, termed as EEG data matrix. However, the resultant EEG data matrix had high dimensionality. Therefore, dimension reduction was performed based on a rank-based feature selection method according to a criterion, i.e., receiver operating characteristic (ROC). As a result, the most significant features were identified and further be utilized during the training and testing of a classification model, i.e., the logistic regression (LR) classifier. Finally, the LR model was validated with 100 iterations of 10-fold cross-validation (10-CV). The classification results were compared with short-time Fourier transform (STFT) analysis, and empirical mode decompositions (EMD). The wavelet features extracted from frontal and temporal EEG data were found statistically significant. In comparison with other time-frequency approaches such as the STFT and EMD, the WT analysis has shown highest classification accuracy, i.e., accuracy = 87.5%, sensitivity = 95%, and specificity = 80%. In conclusion, significant wavelet coefficients extracted from frontal and temporal pre-treatment EEG data involving delta and theta frequency bands may predict antidepressant’s treatment outcome for the MDD patients. PMID:28152063

Diagnosis of breast masses from dynamic contrast-enhanced and diffusion-weighted MR: a machine learning approach.

PubMed

Cai, Hongmin; Peng, Yanxia; Ou, Caiwen; Chen, Minsheng; Li, Li

2014-01-01

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is increasingly used for breast cancer diagnosis as supplementary to conventional imaging techniques. Combining of diffusion-weighted imaging (DWI) of morphology and kinetic features from DCE-MRI to improve the discrimination power of malignant from benign breast masses is rarely reported. The study comprised of 234 female patients with 85 benign and 149 malignant lesions. Four distinct groups of features, coupling with pathological tests, were estimated to comprehensively characterize the pictorial properties of each lesion, which was obtained by a semi-automated segmentation method. Classical machine learning scheme including feature subset selection and various classification schemes were employed to build prognostic model, which served as a foundation for evaluating the combined effects of the multi-sided features for predicting of the types of lesions. Various measurements including cross validation and receiver operating characteristics were used to quantify the diagnostic performances of each feature as well as their combination. Seven features were all found to be statistically different between the malignant and the benign groups and their combination has achieved the highest classification accuracy. The seven features include one pathological variable of age, one morphological variable of slope, three texture features of entropy, inverse difference and information correlation, one kinetic feature of SER and one DWI feature of apparent diffusion coefficient (ADC). Together with the selected diagnostic features, various classical classification schemes were used to test their discrimination power through cross validation scheme. The averaged measurements of sensitivity, specificity, AUC and accuracy are 0.85, 0.89, 90.9% and 0.93, respectively. Multi-sided variables which characterize the morphological, kinetic, pathological properties and DWI measurement of ADC can dramatically improve the discriminatory power of breast lesions.

Semi-supervised classification tool for DubaiSat-2 multispectral imagery

NASA Astrophysics Data System (ADS)

Al-Mansoori, Saeed

2015-10-01

This paper addresses a semi-supervised classification tool based on a pixel-based approach of the multi-spectral satellite imagery. There are not many studies demonstrating such algorithm for the multispectral images, especially when the image consists of 4 bands (Red, Green, Blue and Near Infrared) as in DubaiSat-2 satellite images. The proposed approach utilizes both unsupervised and supervised classification schemes sequentially to identify four classes in the image, namely, water bodies, vegetation, land (developed and undeveloped areas) and paved areas (i.e. roads). The unsupervised classification concept is applied to identify two classes; water bodies and vegetation, based on a well-known index that uses the distinct wavelengths of visible and near-infrared sunlight that is absorbed and reflected by the plants to identify the classes; this index parameter is called "Normalized Difference Vegetation Index (NDVI)". Afterward, the supervised classification is performed by selecting training homogenous samples for roads and land areas. Here, a precise selection of training samples plays a vital role in the classification accuracy. Post classification is finally performed to enhance the classification accuracy, where the classified image is sieved, clumped and filtered before producing final output. Overall, the supervised classification approach produced higher accuracy than the unsupervised method. This paper shows some current preliminary research results which point out the effectiveness of the proposed technique in a virtual perspective.

Classification Consistency and Accuracy for Complex Assessments Using Item Response Theory

ERIC Educational Resources Information Center

Lee, Won-Chan

2010-01-01

In this article, procedures are described for estimating single-administration classification consistency and accuracy indices for complex assessments using item response theory (IRT). This IRT approach was applied to real test data comprising dichotomous and polytomous items. Several different IRT model combinations were considered. Comparisons…

Conceptual Scoring and Classification Accuracy of Vocabulary Testing in Bilingual Children

ERIC Educational Resources Information Center

Anaya, Jissel B.; Peña, Elizabeth D.; Bedore, Lisa M.

2018-01-01

Purpose: This study examined the effects of single-language and conceptual scoring on the vocabulary performance of bilingual children with and without specific language impairment. We assessed classification accuracy across 3 scoring methods. Method: Participants included Spanish-English bilingual children (N = 247) aged 5;1 (years;months) to…

Classification with spatio-temporal interpixel class dependency contexts

NASA Technical Reports Server (NTRS)

Jeon, Byeungwoo; Landgrebe, David A.

1992-01-01

A contextual classifier which can utilize both spatial and temporal interpixel dependency contexts is investigated. After spatial and temporal neighbors are defined, a general form of maximum a posterior spatiotemporal contextual classifier is derived. This contextual classifier is simplified under several assumptions. Joint prior probabilities of the classes of each pixel and its spatial neighbors are modeled by the Gibbs random field. The classification is performed in a recursive manner to allow a computationally efficient contextual classification. Experimental results with bitemporal TM data show significant improvement of classification accuracy over noncontextual pixelwise classifiers. This spatiotemporal contextual classifier should find use in many applications of remote sensing, especially when the classification accuracy is important.

Gastric precancerous diseases classification using CNN with a concise model.

PubMed

Zhang, Xu; Hu, Weiling; Chen, Fei; Liu, Jiquan; Yang, Yuanhang; Wang, Liangjing; Duan, Huilong; Si, Jianmin

2017-01-01

Gastric precancerous diseases (GPD) may deteriorate into early gastric cancer if misdiagnosed, so it is important to help doctors recognize GPD accurately and quickly. In this paper, we realize the classification of 3-class GPD, namely, polyp, erosion, and ulcer using convolutional neural networks (CNN) with a concise model called the Gastric Precancerous Disease Network (GPDNet). GPDNet introduces fire modules from SqueezeNet to reduce the model size and parameters about 10 times while improving speed for quick classification. To maintain classification accuracy with fewer parameters, we propose an innovative method called iterative reinforced learning (IRL). After training GPDNet from scratch, we apply IRL to fine-tune the parameters whose values are close to 0, and then we take the modified model as a pretrained model for the next training. The result shows that IRL can improve the accuracy about 9% after 6 iterations. The final classification accuracy of our GPDNet was 88.90%, which is promising for clinical GPD recognition.

Convolutional neural network with transfer learning for rice type classification

NASA Astrophysics Data System (ADS)

Patel, Vaibhav Amit; Joshi, Manjunath V.

2018-04-01

Presently, rice type is identified manually by humans, which is time consuming and error prone. Therefore, there is a need to do this by machine which makes it faster with greater accuracy. This paper proposes a deep learning based method for classification of rice types. We propose two methods to classify the rice types. In the first method, we train a deep convolutional neural network (CNN) using the given segmented rice images. In the second method, we train a combination of a pretrained VGG16 network and the proposed method, while using transfer learning in which the weights of a pretrained network are used to achieve better accuracy. Our approach can also be used for classification of rice grain as broken or fine. We train a 5-class model for classifying rice types using 4000 training images and another 2- class model for the classification of broken and normal rice using 1600 training images. We observe that despite having distinct rice images, our architecture, pretrained on ImageNet data boosts classification accuracy significantly.

Evaluating the High Risk Groups for Suicide: A Comparison of Logistic Regression, Support Vector Machine, Decision Tree and Artificial Neural Network

PubMed Central

AMINI, Payam; AHMADINIA, Hasan; POOROLAJAL, Jalal; MOQADDASI AMIRI, Mohammad

2016-01-01

Background: We aimed to assess the high-risk group for suicide using different classification methods includinglogistic regression (LR), decision tree (DT), artificial neural network (ANN), and support vector machine (SVM). Methods: We used the dataset of a study conducted to predict risk factors of completed suicide in Hamadan Province, the west of Iran, in 2010. To evaluate the high-risk groups for suicide, LR, SVM, DT and ANN were performed. The applied methods were compared using sensitivity, specificity, positive predicted value, negative predicted value, accuracy and the area under curve. Cochran-Q test was implied to check differences in proportion among methods. To assess the association between the observed and predicted values, Ø coefficient, contingency coefficient, and Kendall tau-b were calculated. Results: Gender, age, and job were the most important risk factors for fatal suicide attempts in common for four methods. SVM method showed the highest accuracy 0.68 and 0.67 for training and testing sample, respectively. However, this method resulted in the highest specificity (0.67 for training and 0.68 for testing sample) and the highest sensitivity for training sample (0.85), but the lowest sensitivity for the testing sample (0.53). Cochran-Q test resulted in differences between proportions in different methods (P<0.001). The association of SVM predictions and observed values, Ø coefficient, contingency coefficient, and Kendall tau-b were 0.239, 0.232 and 0.239, respectively. Conclusion: SVM had the best performance to classify fatal suicide attempts comparing to DT, LR and ANN. PMID:27957463

PSO/ACO algorithm-based risk assessment of human neural tube defects in Heshun County, China.

PubMed

Liao, Yi Lan; Wang, Jin Feng; Wu, Ji Lei; Wang, Jiao Jiao; Zheng, Xiao Ying

2012-10-01

To develop a new technique for assessing the risk of birth defects, which are a major cause of infant mortality and disability in many parts of the world. The region of interest in this study was Heshun County, the county in China with the highest rate of neural tube defects (NTDs). A hybrid particle swarm optimization/ant colony optimization (PSO/ACO) algorithm was used to quantify the probability of NTDs occurring at villages with no births. The hybrid PSO/ACO algorithm is a form of artificial intelligence adapted for hierarchical classification. It is a powerful technique for modeling complex problems involving impacts of causes. The algorithm was easy to apply, with the accuracy of the results being 69.5%±7.02% at the 95% confidence level. The proposed method is simple to apply, has acceptable fault tolerance, and greatly enhances the accuracy of calculations. Copyright © 2012 The Editorial Board of Biomedical and Environmental Sciences. Published by Elsevier B.V. All rights reserved.

Classification systems for lower extremity amputation prediction in subjects with active diabetic foot ulcer: a systematic review and meta-analysis.

PubMed

Monteiro-Soares, M; Martins-Mendes, D; Vaz-Carneiro, A; Sampaio, S; Dinis-Ribeiro, M

2014-10-01

We systematically review the available systems used to classify diabetic foot ulcers in order to synthesize their methodological qualitative issues and accuracy to predict lower extremity amputation, as this may represent a critical point in these patients' care. Two investigators searched, in EBSCO, ISI, PubMed and SCOPUS databases, and independently selected studies published until May 2013 and reporting prognostic accuracy and/or reliability of specific systems for patients with diabetic foot ulcer in order to predict lower extremity amputation. We included 25 studies reporting a prevalence of lower extremity amputation between 6% and 78%. Eight different diabetic foot ulcer descriptions and seven prognostic stratification classification systems were addressed with a variable (1-9) number of factors included, specially peripheral arterial disease (n = 12) or infection at the ulcer site (n = 10) or ulcer depth (n = 10). The Meggitt-Wagner, S(AD)SAD and Texas University Classification systems were the most extensively validated, whereas ten classifications were derived or validated only once. Reliability was reported in a single study, and accuracy measures were reported in five studies with another eight allowing their calculation. Pooled accuracy ranged from 0.65 (for gangrene) to 0.74 (for infection). There are numerous classification systems for diabetic foot ulcer outcome prediction, but only few studies evaluated their reliability or external validity. Studies rarely validated several systems simultaneously and only a few reported accuracy measures. Further studies assessing reliability and accuracy of the available systems and their composing variables are needed. Copyright © 2014 John Wiley & Sons, Ltd.

Classification of EMG signals using PSO optimized SVM for diagnosis of neuromuscular disorders.

PubMed

Subasi, Abdulhamit

2013-06-01

Support vector machine (SVM) is an extensively used machine learning method with many biomedical signal classification applications. In this study, a novel PSO-SVM model has been proposed that hybridized the particle swarm optimization (PSO) and SVM to improve the EMG signal classification accuracy. This optimization mechanism involves kernel parameter setting in the SVM training procedure, which significantly influences the classification accuracy. The experiments were conducted on the basis of EMG signal to classify into normal, neurogenic or myopathic. In the proposed method the EMG signals were decomposed into the frequency sub-bands using discrete wavelet transform (DWT) and a set of statistical features were extracted from these sub-bands to represent the distribution of wavelet coefficients. The obtained results obviously validate the superiority of the SVM method compared to conventional machine learning methods, and suggest that further significant enhancements in terms of classification accuracy can be achieved by the proposed PSO-SVM classification system. The PSO-SVM yielded an overall accuracy of 97.41% on 1200 EMG signals selected from 27 subject records against 96.75%, 95.17% and 94.08% for the SVM, the k-NN and the RBF classifiers, respectively. PSO-SVM is developed as an efficient tool so that various SVMs can be used conveniently as the core of PSO-SVM for diagnosis of neuromuscular disorders. Copyright © 2013 Elsevier Ltd. All rights reserved.

Accuracy of parasitological and immunological tests for the screening of human schistosomiasis in immigrants and refugees from African countries: An approach with Latent Class Analysis

PubMed Central

Beltrame, Anna; Guerriero, Massimo; Angheben, Andrea; Gobbi, Federico; Requena-Mendez, Ana; Zammarchi, Lorenzo; Formenti, Fabio; Perandin, Francesca; Bisoffi, Zeno

2017-01-01

Background Schistosomiasis is a neglected infection affecting millions of people, mostly living in sub-Saharan Africa. Morbidity and mortality due to chronic infection are relevant, although schistosomiasis is often clinically silent. Different diagnostic tests have been implemented in order to improve screening and diagnosis, that traditionally rely on parasitological tests with low sensitivity. Aim of this study was to evaluate the accuracy of different tests for the screening of schistosomiasis in African migrants, in a non endemic setting. Methodology/Principal findings A retrospective study was conducted on 373 patients screened at the Centre for Tropical Diseases (CTD) in Negrar, Verona, Italy. Biological samples were tested with: stool/urine microscopy, Circulating Cathodic Antigen (CCA) dipstick test, ELISA, Western blot, immune-chromatographic test (ICT). Test accuracy and predictive values of the immunological tests were assessed primarily on the basis of the results of microscopy (primary reference standard): ICT and WB resulted the test with highest sensitivity (94% and 92%, respectively), with a high NPV (98%). CCA showed the highest specificity (93%), but low sensitivity (48%). The analysis was conducted also using a composite reference standard, CRS (patients classified as infected in case of positive microscopy and/or at least 2 concordant positive immunological tests) and Latent Class Analysis (LCA). The latter two models demonstrated excellent agreement (Cohen’s kappa: 0.92) for the classification of the results. In fact, they both confirmed ICT as the test with the highest sensitivity (96%) and NPV (97%), moreover PPV was reasonably good (78% and 72% according to CRS and LCA, respectively). ELISA resulted the most specific immunological test (over 99%). The ICT appears to be a suitable screening test, even when used alone. Conclusions The rapid test ICT was the most sensitive test, with the potential of being used as a single screening test for African migrants. PMID:28582412

Urban Image Classification: Per-Pixel Classifiers, Sub-Pixel Analysis, Object-Based Image Analysis, and Geospatial Methods. 10; Chapter

NASA Technical Reports Server (NTRS)

Myint, Soe W.; Mesev, Victor; Quattrochi, Dale; Wentz, Elizabeth A.

2013-01-01

Remote sensing methods used to generate base maps to analyze the urban environment rely predominantly on digital sensor data from space-borne platforms. This is due in part from new sources of high spatial resolution data covering the globe, a variety of multispectral and multitemporal sources, sophisticated statistical and geospatial methods, and compatibility with GIS data sources and methods. The goal of this chapter is to review the four groups of classification methods for digital sensor data from space-borne platforms; per-pixel, sub-pixel, object-based (spatial-based), and geospatial methods. Per-pixel methods are widely used methods that classify pixels into distinct categories based solely on the spectral and ancillary information within that pixel. They are used for simple calculations of environmental indices (e.g., NDVI) to sophisticated expert systems to assign urban land covers. Researchers recognize however, that even with the smallest pixel size the spectral information within a pixel is really a combination of multiple urban surfaces. Sub-pixel classification methods therefore aim to statistically quantify the mixture of surfaces to improve overall classification accuracy. While within pixel variations exist, there is also significant evidence that groups of nearby pixels have similar spectral information and therefore belong to the same classification category. Object-oriented methods have emerged that group pixels prior to classification based on spectral similarity and spatial proximity. Classification accuracy using object-based methods show significant success and promise for numerous urban 3 applications. Like the object-oriented methods that recognize the importance of spatial proximity, geospatial methods for urban mapping also utilize neighboring pixels in the classification process. The primary difference though is that geostatistical methods (e.g., spatial autocorrelation methods) are utilized during both the pre- and post-classification steps. Within this chapter, each of the four approaches is described in terms of scale and accuracy classifying urban land use and urban land cover; and for its range of urban applications. We demonstrate the overview of four main classification groups in Figure 1 while Table 1 details the approaches with respect to classification requirements and procedures (e.g., reflectance conversion, steps before training sample selection, training samples, spatial approaches commonly used, classifiers, primary inputs for classification, output structures, number of output layers, and accuracy assessment). The chapter concludes with a brief summary of the methods reviewed and the challenges that remain in developing new classification methods for improving the efficiency and accuracy of mapping urban areas.

The Upper and Lower Bounds of the Prediction Accuracies of Ensemble Methods for Binary Classification

PubMed Central

Wang, Xueyi; Davidson, Nicholas J.

2011-01-01

Ensemble methods have been widely used to improve prediction accuracy over individual classifiers. In this paper, we achieve a few results about the prediction accuracies of ensemble methods for binary classification that are missed or misinterpreted in previous literature. First we show the upper and lower bounds of the prediction accuracies (i.e. the best and worst possible prediction accuracies) of ensemble methods. Next we show that an ensemble method can achieve > 0.5 prediction accuracy, while individual classifiers have < 0.5 prediction accuracies. Furthermore, for individual classifiers with different prediction accuracies, the average of the individual accuracies determines the upper and lower bounds. We perform two experiments to verify the results and show that it is hard to achieve the upper and lower bounds accuracies by random individual classifiers and better algorithms need to be developed. PMID:21853162

Mapping regional distribution of a single tree species: Whitebark pine in the Greater Yellowstone Ecosystem

USGS Publications Warehouse

Landenburger, L.; Lawrence, R.L.; Podruzny, S.; Schwartz, C.C.

2008-01-01

Moderate resolution satellite imagery traditionally has been thought to be inadequate for mapping vegetation at the species level. This has made comprehensive mapping of regional distributions of sensitive species, such as whitebark pine, either impractical or extremely time consuming. We sought to determine whether using a combination of moderate resolution satellite imagery (Landsat Enhanced Thematic Mapper Plus), extensive stand data collected by land management agencies for other purposes, and modern statistical classification techniques (boosted classification trees) could result in successful mapping of whitebark pine. Overall classification accuracies exceeded 90%, with similar individual class accuracies. Accuracies on a localized basis varied based on elevation. Accuracies also varied among administrative units, although we were not able to determine whether these differences related to inherent spatial variations or differences in the quality of available reference data.

The effect of spatial, spectral and radiometric factors on classification accuracy using thematic mapper data

NASA Technical Reports Server (NTRS)

Wrigley, R. C.; Acevedo, W.; Alexander, D.; Buis, J.; Card, D.

1984-01-01

An experiment of a factorial design was conducted to test the effects on classification accuracy of land cover types due to the improved spatial, spectral and radiometric characteristics of the Thematic Mapper (TM) in comparison to the Multispectral Scanner (MSS). High altitude aircraft scanner data from the Airborne Thematic Mapper instrument was acquired over central California in August, 1983 and used to simulate Thematic Mapper data as well as all combinations of the three characteristics for eight data sets in all. Results for the training sites (field center pixels) showed better classification accuracies for MSS spatial resolution, TM spectral bands and TM radiometry in order of importance.

Combined data mining/NIR spectroscopy for purity assessment of lime juice

NASA Astrophysics Data System (ADS)

Shafiee, Sahameh; Minaei, Saeid

2018-06-01

This paper reports the data mining study on the NIR spectrum of lime juice samples to determine their purity (natural or synthetic). NIR spectra for 72 pure and synthetic lime juice samples were recorded in reflectance mode. Sample outliers were removed using PCA analysis. Different data mining techniques for feature selection (Genetic Algorithm (GA)) and classification (including the radial basis function (RBF) network, Support Vector Machine (SVM), and Random Forest (RF) tree) were employed. Based on the results, SVM proved to be the most accurate classifier as it achieved the highest accuracy (97%) using the raw spectrum information. The classifier accuracy dropped to 93% when selected feature vector by GA search method was applied as classifier input. It can be concluded that some relevant features which produce good performance with the SVM classifier are removed by feature selection. Also, reduced spectra using PCA do not show acceptable performance (total accuracy of 66% by RBFNN), which indicates that dimensional reduction methods such as PCA do not always lead to more accurate results. These findings demonstrate the potential of data mining combination with near-infrared spectroscopy for monitoring lime juice quality in terms of natural or synthetic nature.

MKID digital readout tuning with deep learning

NASA Astrophysics Data System (ADS)

Dodkins, R.; Mahashabde, S.; O'Brien, K.; Thatte, N.; Fruitwala, N.; Walter, A. B.; Meeker, S. R.; Szypryt, P.; Mazin, B. A.

2018-04-01

Microwave Kinetic Inductance Detector (MKID) devices offer inherent spectral resolution, simultaneous read out of thousands of pixels, and photon-limited sensitivity at optical wavelengths. Before taking observations the readout power and frequency of each pixel must be individually tuned, and if the equilibrium state of the pixels change, then the readout must be retuned. This process has previously been performed through manual inspection, and typically takes one hour per 500 resonators (20 h for a ten-kilo-pixel array). We present an algorithm based on a deep convolution neural network (CNN) architecture to determine the optimal bias power for each resonator. The bias point classifications from this CNN model, and those from alternative automated methods, are compared to those from human decisions, and the accuracy of each method is assessed. On a test feed-line dataset, the CNN achieves an accuracy of 90% within 1 dB of the designated optimal value, which is equivalent accuracy to a randomly selected human operator, and superior to the highest scoring alternative automated method by 10%. On a full ten-kilopixel array, the CNN performs the characterization in a matter of minutes - paving the way for future mega-pixel MKID arrays.

Automated Detection of P. falciparum Using Machine Learning Algorithms with Quantitative Phase Images of Unstained Cells.

PubMed

Park, Han Sang; Rinehart, Matthew T; Walzer, Katelyn A; Chi, Jen-Tsan Ashley; Wax, Adam

2016-01-01

Malaria detection through microscopic examination of stained blood smears is a diagnostic challenge that heavily relies on the expertise of trained microscopists. This paper presents an automated analysis method for detection and staging of red blood cells infected by the malaria parasite Plasmodium falciparum at trophozoite or schizont stage. Unlike previous efforts in this area, this study uses quantitative phase images of unstained cells. Erythrocytes are automatically segmented using thresholds of optical phase and refocused to enable quantitative comparison of phase images. Refocused images are analyzed to extract 23 morphological descriptors based on the phase information. While all individual descriptors are highly statistically different between infected and uninfected cells, each descriptor does not enable separation of populations at a level satisfactory for clinical utility. To improve the diagnostic capacity, we applied various machine learning techniques, including linear discriminant classification (LDC), logistic regression (LR), and k-nearest neighbor classification (NNC), to formulate algorithms that combine all of the calculated physical parameters to distinguish cells more effectively. Results show that LDC provides the highest accuracy of up to 99.7% in detecting schizont stage infected cells compared to uninfected RBCs. NNC showed slightly better accuracy (99.5%) than either LDC (99.0%) or LR (99.1%) for discriminating late trophozoites from uninfected RBCs. However, for early trophozoites, LDC produced the best accuracy of 98%. Discrimination of infection stage was less accurate, producing high specificity (99.8%) but only 45.0%-66.8% sensitivity with early trophozoites most often mistaken for late trophozoite or schizont stage and late trophozoite and schizont stage most often confused for each other. Overall, this methodology points to a significant clinical potential of using quantitative phase imaging to detect and stage malaria infection without staining or expert analysis.

Automated Detection of P. falciparum Using Machine Learning Algorithms with Quantitative Phase Images of Unstained Cells

PubMed Central

Park, Han Sang; Rinehart, Matthew T.; Walzer, Katelyn A.; Chi, Jen-Tsan Ashley; Wax, Adam

2016-01-01

Malaria detection through microscopic examination of stained blood smears is a diagnostic challenge that heavily relies on the expertise of trained microscopists. This paper presents an automated analysis method for detection and staging of red blood cells infected by the malaria parasite Plasmodium falciparum at trophozoite or schizont stage. Unlike previous efforts in this area, this study uses quantitative phase images of unstained cells. Erythrocytes are automatically segmented using thresholds of optical phase and refocused to enable quantitative comparison of phase images. Refocused images are analyzed to extract 23 morphological descriptors based on the phase information. While all individual descriptors are highly statistically different between infected and uninfected cells, each descriptor does not enable separation of populations at a level satisfactory for clinical utility. To improve the diagnostic capacity, we applied various machine learning techniques, including linear discriminant classification (LDC), logistic regression (LR), and k-nearest neighbor classification (NNC), to formulate algorithms that combine all of the calculated physical parameters to distinguish cells more effectively. Results show that LDC provides the highest accuracy of up to 99.7% in detecting schizont stage infected cells compared to uninfected RBCs. NNC showed slightly better accuracy (99.5%) than either LDC (99.0%) or LR (99.1%) for discriminating late trophozoites from uninfected RBCs. However, for early trophozoites, LDC produced the best accuracy of 98%. Discrimination of infection stage was less accurate, producing high specificity (99.8%) but only 45.0%-66.8% sensitivity with early trophozoites most often mistaken for late trophozoite or schizont stage and late trophozoite and schizont stage most often confused for each other. Overall, this methodology points to a significant clinical potential of using quantitative phase imaging to detect and stage malaria infection without staining or expert analysis. PMID:27636719

A study of the effectiveness of machine learning methods for classification of clinical interview fragments into a large number of categories.

PubMed

Hasan, Mehedi; Kotov, Alexander; Carcone, April; Dong, Ming; Naar, Sylvie; Hartlieb, Kathryn Brogan

2016-08-01

This study examines the effectiveness of state-of-the-art supervised machine learning methods in conjunction with different feature types for the task of automatic annotation of fragments of clinical text based on codebooks with a large number of categories. We used a collection of motivational interview transcripts consisting of 11,353 utterances, which were manually annotated by two human coders as the gold standard, and experimented with state-of-art classifiers, including Naïve Bayes, J48 Decision Tree, Support Vector Machine (SVM), Random Forest (RF), AdaBoost, DiscLDA, Conditional Random Fields (CRF) and Convolutional Neural Network (CNN) in conjunction with lexical, contextual (label of the previous utterance) and semantic (distribution of words in the utterance across the Linguistic Inquiry and Word Count dictionaries) features. We found out that, when the number of classes is large, the performance of CNN and CRF is inferior to SVM. When only lexical features were used, interview transcripts were automatically annotated by SVM with the highest classification accuracy among all classifiers of 70.8%, 61% and 53.7% based on the codebooks consisting of 17, 20 and 41 codes, respectively. Using contextual and semantic features, as well as their combination, in addition to lexical ones, improved the accuracy of SVM for annotation of utterances in motivational interview transcripts with a codebook consisting of 17 classes to 71.5%, 74.2%, and 75.1%, respectively. Our results demonstrate the potential of using machine learning methods in conjunction with lexical, semantic and contextual features for automatic annotation of clinical interview transcripts with near-human accuracy. Copyright © 2016 Elsevier Inc. All rights reserved.

Multi-temporal Land Use Mapping of Coastal Wetlands Area using Machine Learning in Google Earth Engine

NASA Astrophysics Data System (ADS)

Farda, N. M.

2017-12-01

Coastal wetlands provide ecosystem services essential to people and the environment. Changes in coastal wetlands, especially on land use, are important to monitor by utilizing multi-temporal imagery. The Google Earth Engine (GEE) provides many machine learning algorithms (10 algorithms) that are very useful for extracting land use from imagery. The research objective is to explore machine learning in Google Earth Engine and its accuracy for multi-temporal land use mapping of coastal wetland area. Landsat 3 MSS (1978), Landsat 5 TM (1991), Landsat 7 ETM+ (2001), and Landsat 8 OLI (2014) images located in Segara Anakan lagoon are selected to represent multi temporal images. The input for machine learning are visible and near infrared bands, PCA band, invers PCA bands, bare soil index, vegetation index, wetness index, elevation from ASTER GDEM, and GLCM (Harralick) texture, and also polygon samples in 140 locations. There are 10 machine learning algorithms applied to extract coastal wetlands land use from Landsat imagery. The algorithms are Fast Naive Bayes, CART (Classification and Regression Tree), Random Forests, GMO Max Entropy, Perceptron (Multi Class Perceptron), Winnow, Voting SVM, Margin SVM, Pegasos (Primal Estimated sub-GrAdient SOlver for Svm), IKPamir (Intersection Kernel Passive Aggressive Method for Information Retrieval, SVM). Machine learning in Google Earth Engine are very helpful in multi-temporal land use mapping, the highest accuracy for land use mapping of coastal wetland is CART with 96.98 % Overall Accuracy using K-Fold Cross Validation (K = 10). GEE is particularly useful for multi-temporal land use mapping with ready used image and classification algorithms, and also very challenging for other applications.

Multiple confidence estimates as indices of eyewitness memory.

PubMed

Sauer, James D; Brewer, Neil; Weber, Nathan

2008-08-01

Eyewitness identification decisions are vulnerable to various influences on witnesses' decision criteria that contribute to false identifications of innocent suspects and failures to choose perpetrators. An alternative procedure using confidence estimates to assess the degree of match between novel and previously viewed faces was investigated. Classification algorithms were applied to participants' confidence data to determine when a confidence value or pattern of confidence values indicated a positive response. Experiment 1 compared confidence group classification accuracy with a binary decision control group's accuracy on a standard old-new face recognition task and found superior accuracy for the confidence group for target-absent trials but not for target-present trials. Experiment 2 used a face mini-lineup task and found reduced target-present accuracy offset by large gains in target-absent accuracy. Using a standard lineup paradigm, Experiments 3 and 4 also found improved classification accuracy for target-absent lineups and, with a more sophisticated algorithm, for target-present lineups. This demonstrates the accessibility of evidence for recognition memory decisions and points to a more sensitive index of memory quality than is afforded by binary decisions.

Vehicle Classification Using an Imbalanced Dataset Based on a Single Magnetic Sensor.

PubMed

Xu, Chang; Wang, Yingguan; Bao, Xinghe; Li, Fengrong

2018-05-24

This paper aims to improve the accuracy of automatic vehicle classifiers for imbalanced datasets. Classification is made through utilizing a single anisotropic magnetoresistive sensor, with the models of vehicles involved being classified into hatchbacks, sedans, buses, and multi-purpose vehicles (MPVs). Using time domain and frequency domain features in combination with three common classification algorithms in pattern recognition, we develop a novel feature extraction method for vehicle classification. These three common classification algorithms are the k-nearest neighbor, the support vector machine, and the back-propagation neural network. Nevertheless, a problem remains with the original vehicle magnetic dataset collected being imbalanced, and may lead to inaccurate classification results. With this in mind, we propose an approach called SMOTE, which can further boost the performance of classifiers. Experimental results show that the k-nearest neighbor (KNN) classifier with the SMOTE algorithm can reach a classification accuracy of 95.46%, thus minimizing the effect of the imbalance.

Evaluation of multiband, multitemporal, and transformed LANDSAT MSS data for land cover area estimation. [North Central Missouri

NASA Technical Reports Server (NTRS)

Stoner, E. R.; May, G. A.; Kalcic, M. T. (Principal Investigator)

1981-01-01

Sample segments of ground-verified land cover data collected in conjunction with the USDA/ESS June Enumerative Survey were merged with LANDSAT data and served as a focus for unsupervised spectral class development and accuracy assessment. Multitemporal data sets were created from single-date LANDSAT MSS acquisitions from a nominal scene covering an eleven-county area in north central Missouri. Classification accuracies for the four land cover types predominant in the test site showed significant improvement in going from unitemporal to multitemporal data sets. Transformed LANDSAT data sets did not significantly improve classification accuracies. Regression estimators yielded mixed results for different land covers. Misregistration of two LANDSAT data sets by as much and one half pixels did not significantly alter overall classification accuracies. Existing algorithms for scene-to scene overlay proved adequate for multitemporal data analysis as long as statistical class development and accuracy assessment were restricted to field interior pixels.

Examining the Classification Accuracy of a Vocabulary Screening Measure with Preschool Children

ERIC Educational Resources Information Center

Marcotte, Amanda M.; Clemens, Nathan H.; Parker, Christopher; Whitcomb, Sara A.

2016-01-01

This study investigated the classification accuracy of the "Dynamic Indicators of Vocabulary Skills" (DIVS) as a preschool vocabulary screening measure. With a sample of 240 preschoolers, fall and winter DIVS scores were used to predict year-end vocabulary risk using the 25th percentile on the "Peabody Picture Vocabulary Test--Third…

Classification Accuracy and Acceptability of the Integrated Screening and Intervention System Teacher Rating Form

ERIC Educational Resources Information Center

Daniels, Brian; Volpe, Robert J.; Fabiano, Gregory A.; Briesch, Amy M.

2017-01-01

This study examines the classification accuracy and teacher acceptability of a problem-focused screener for academic and disruptive behavior problems, which is directly linked to evidence-based intervention. Participants included 39 classroom teachers from 2 public school districts in the Northeastern United States. Teacher ratings were obtained…

Assessing the Accuracy and Consistency of Language Proficiency Classification under Competing Measurement Models

ERIC Educational Resources Information Center

Zhang, Bo

2010-01-01

This article investigates how measurement models and statistical procedures can be applied to estimate the accuracy of proficiency classification in language testing. The paper starts with a concise introduction of four measurement models: the classical test theory (CTT) model, the dichotomous item response theory (IRT) model, the testlet response…

Concurrent Validity and Classification Accuracy of Curriculum-Based Measurement for Written Expression

ERIC Educational Resources Information Center

Furey, William M.; Marcotte, Amanda M.; Hintze, John M.; Shackett, Caroline M.

2016-01-01

The study presents a critical analysis of written expression curriculum-based measurement (WE-CBM) metrics derived from 3- and 10-min test lengths. Criterion validity and classification accuracy were examined for Total Words Written (TWW), Correct Writing Sequences (CWS), Percent Correct Writing Sequences (%CWS), and Correct Minus Incorrect…

Dynamic Assessment of School-Age Children's Narrative Ability: An Experimental Investigation of Classification Accuracy

ERIC Educational Resources Information Center

Pena, Elizabeth D.; Gillam, Ronald B.; Malek, Melynn; Ruiz-Felter, Roxanna; Resendiz, Maria; Fiestas, Christine; Sabel, Tracy

2006-01-01

Two experiments examined reliability and classification accuracy of a narration-based dynamic assessment task. Purpose: The first experiment evaluated whether parallel results were obtained from stories created in response to 2 different wordless picture books. If so, the tasks and measures would be appropriate for assessing pretest and posttest…

The Potential Impact of Not Being Able to Create Parallel Tests on Expected Classification Accuracy

ERIC Educational Resources Information Center

Wyse, Adam E.

2011-01-01

In many practical testing situations, alternate test forms from the same testing program are not strictly parallel to each other and instead the test forms exhibit small psychometric differences. This article investigates the potential practical impact that these small psychometric differences can have on expected classification accuracy. Ten…

Emotion recognition from multichannel EEG signals using K-nearest neighbor classification.

PubMed

Li, Mi; Xu, Hongpei; Liu, Xingwang; Lu, Shengfu

2018-04-27

Many studies have been done on the emotion recognition based on multi-channel electroencephalogram (EEG) signals. This paper explores the influence of the emotion recognition accuracy of EEG signals in different frequency bands and different number of channels. We classified the emotional states in the valence and arousal dimensions using different combinations of EEG channels. Firstly, DEAP default preprocessed data were normalized. Next, EEG signals were divided into four frequency bands using discrete wavelet transform, and entropy and energy were calculated as features of K-nearest neighbor Classifier. The classification accuracies of the 10, 14, 18 and 32 EEG channels based on the Gamma frequency band were 89.54%, 92.28%, 93.72% and 95.70% in the valence dimension and 89.81%, 92.24%, 93.69% and 95.69% in the arousal dimension. As the number of channels increases, the classification accuracy of emotional states also increases, the classification accuracy of the gamma frequency band is greater than that of the beta frequency band followed by the alpha and theta frequency bands. This paper provided better frequency bands and channels reference for emotion recognition based on EEG.

Gender classification in children based on speech characteristics: using fundamental and formant frequencies of Malay vowels.

PubMed

Zourmand, Alireza; Ting, Hua-Nong; Mirhassani, Seyed Mostafa

2013-03-01

Speech is one of the prevalent communication mediums for humans. Identifying the gender of a child speaker based on his/her speech is crucial in telecommunication and speech therapy. This article investigates the use of fundamental and formant frequencies from sustained vowel phonation to distinguish the gender of Malay children aged between 7 and 12 years. The Euclidean minimum distance and multilayer perceptron were used to classify the gender of 360 Malay children based on different combinations of fundamental and formant frequencies (F0, F1, F2, and F3). The Euclidean minimum distance with normalized frequency data achieved a classification accuracy of 79.44%, which was higher than that of the nonnormalized frequency data. Age-dependent modeling was used to improve the accuracy of gender classification. The Euclidean distance method obtained 84.17% based on the optimal classification accuracy for all age groups. The accuracy was further increased to 99.81% using multilayer perceptron based on mel-frequency cepstral coefficients. Copyright © 2013 The Voice Foundation. Published by Mosby, Inc. All rights reserved.

Classification of pulmonary pathology from breath sounds using the wavelet packet transform and an extreme learning machine.

PubMed

Palaniappan, Rajkumar; Sundaraj, Kenneth; Sundaraj, Sebastian; Huliraj, N; Revadi, S S

2017-06-08

Auscultation is a medical procedure used for the initial diagnosis and assessment of lung and heart diseases. From this perspective, we propose assessing the performance of the extreme learning machine (ELM) classifiers for the diagnosis of pulmonary pathology using breath sounds. Energy and entropy features were extracted from the breath sound using the wavelet packet transform. The statistical significance of the extracted features was evaluated by one-way analysis of variance (ANOVA). The extracted features were inputted into the ELM classifier. The maximum classification accuracies obtained for the conventional validation (CV) of the energy and entropy features were 97.36% and 98.37%, respectively, whereas the accuracies obtained for the cross validation (CRV) of the energy and entropy features were 96.80% and 97.91%, respectively. In addition, maximum classification accuracies of 98.25% and 99.25% were obtained for the CV and CRV of the ensemble features, respectively. The results indicate that the classification accuracy obtained with the ensemble features was higher than those obtained with the energy and entropy features.

Impacts of land use/cover classification accuracy on regional climate simulations

NASA Astrophysics Data System (ADS)

Ge, Jianjun; Qi, Jiaguo; Lofgren, Brent M.; Moore, Nathan; Torbick, Nathan; Olson, Jennifer M.

2007-03-01

Land use/cover change has been recognized as a key component in global change. Various land cover data sets, including historically reconstructed, recently observed, and future projected, have been used in numerous climate modeling studies at regional to global scales. However, little attention has been paid to the effect of land cover classification accuracy on climate simulations, though accuracy assessment has become a routine procedure in land cover production community. In this study, we analyzed the behavior of simulated precipitation in the Regional Atmospheric Modeling System (RAMS) over a range of simulated classification accuracies over a 3 month period. This study found that land cover accuracy under 80% had a strong effect on precipitation especially when the land surface had a greater control of the atmosphere. This effect became stronger as the accuracy decreased. As shown in three follow-on experiments, the effect was further influenced by model parameterizations such as convection schemes and interior nudging, which can mitigate the strength of surface boundary forcings. In reality, land cover accuracy rarely obtains the commonly recommended 85% target. Its effect on climate simulations should therefore be considered, especially when historically reconstructed and future projected land covers are employed.

Dimensionality-varied convolutional neural network for spectral-spatial classification of hyperspectral data

NASA Astrophysics Data System (ADS)

Liu, Wanjun; Liang, Xuejian; Qu, Haicheng

2017-11-01

Hyperspectral image (HSI) classification is one of the most popular topics in remote sensing community. Traditional and deep learning-based classification methods were proposed constantly in recent years. In order to improve the classification accuracy and robustness, a dimensionality-varied convolutional neural network (DVCNN) was proposed in this paper. DVCNN was a novel deep architecture based on convolutional neural network (CNN). The input of DVCNN was a set of 3D patches selected from HSI which contained spectral-spatial joint information. In the following feature extraction process, each patch was transformed into some different 1D vectors by 3D convolution kernels, which were able to extract features from spectral-spatial data. The rest of DVCNN was about the same as general CNN and processed 2D matrix which was constituted by by all 1D data. So that the DVCNN could not only extract more accurate and rich features than CNN, but also fused spectral-spatial information to improve classification accuracy. Moreover, the robustness of network on water-absorption bands was enhanced in the process of spectral-spatial fusion by 3D convolution, and the calculation was simplified by dimensionality varied convolution. Experiments were performed on both Indian Pines and Pavia University scene datasets, and the results showed that the classification accuracy of DVCNN improved by 32.87% on Indian Pines and 19.63% on Pavia University scene than spectral-only CNN. The maximum accuracy improvement of DVCNN achievement was 13.72% compared with other state-of-the-art HSI classification methods, and the robustness of DVCNN on water-absorption bands noise was demonstrated.

Integrative Chemical-Biological Read-Across Approach for Chemical Hazard Classification

PubMed Central

Low, Yen; Sedykh, Alexander; Fourches, Denis; Golbraikh, Alexander; Whelan, Maurice; Rusyn, Ivan; Tropsha, Alexander

2013-01-01

Traditional read-across approaches typically rely on the chemical similarity principle to predict chemical toxicity; however, the accuracy of such predictions is often inadequate due to the underlying complex mechanisms of toxicity. Here we report on the development of a hazard classification and visualization method that draws upon both chemical structural similarity and comparisons of biological responses to chemicals measured in multiple short-term assays (”biological” similarity). The Chemical-Biological Read-Across (CBRA) approach infers each compound's toxicity from those of both chemical and biological analogs whose similarities are determined by the Tanimoto coefficient. Classification accuracy of CBRA was compared to that of classical RA and other methods using chemical descriptors alone, or in combination with biological data. Different types of adverse effects (hepatotoxicity, hepatocarcinogenicity, mutagenicity, and acute lethality) were classified using several biological data types (gene expression profiling and cytotoxicity screening). CBRA-based hazard classification exhibited consistently high external classification accuracy and applicability to diverse chemicals. Transparency of the CBRA approach is aided by the use of radial plots that show the relative contribution of analogous chemical and biological neighbors. Identification of both chemical and biological features that give rise to the high accuracy of CBRA-based toxicity prediction facilitates mechanistic interpretation of the models. PMID:23848138

Classification of ECG beats using deep belief network and active learning.

PubMed

G, Sayantan; T, Kien P; V, Kadambari K

2018-04-12

A new semi-supervised approach based on deep learning and active learning for classification of electrocardiogram signals (ECG) is proposed. The objective of the proposed work is to model a scientific method for classification of cardiac irregularities using electrocardiogram beats. The model follows the Association for the Advancement of medical instrumentation (AAMI) standards and consists of three phases. In phase I, feature representation of ECG is learnt using Gaussian-Bernoulli deep belief network followed by a linear support vector machine (SVM) training in the consecutive phase. It yields three deep models which are based on AAMI-defined classes, namely N, V, S, and F. In the last phase, a query generator is introduced to interact with the expert to label few beats to improve accuracy and sensitivity. The proposed approach depicts significant improvement in accuracy with minimal queries posed to the expert and fast online training as tested on the MIT-BIH Arrhythmia Database and the MIT-BIH Supra-ventricular Arrhythmia Database (SVDB). With 100 queries labeled by the expert in phase III, the method achieves an accuracy of 99.5% in "S" versus all classifications (SVEB) and 99.4% accuracy in "V " versus all classifications (VEB) on MIT-BIH Arrhythmia Database. In a similar manner, it is attributed that an accuracy of 97.5% for SVEB and 98.6% for VEB on SVDB database is achieved respectively. Graphical Abstract Reply- Deep belief network augmented by active learning for efficient prediction of arrhythmia.

Study on Classification Accuracy Inspection of Land Cover Data Aided by Automatic Image Change Detection Technology

NASA Astrophysics Data System (ADS)

Xie, W.-J.; Zhang, L.; Chen, H.-P.; Zhou, J.; Mao, W.-J.

2018-04-01

The purpose of carrying out national geographic conditions monitoring is to obtain information of surface changes caused by human social and economic activities, so that the geographic information can be used to offer better services for the government, enterprise and public. Land cover data contains detailed geographic conditions information, thus has been listed as one of the important achievements in the national geographic conditions monitoring project. At present, the main issue of the production of the land cover data is about how to improve the classification accuracy. For the land cover data quality inspection and acceptance, classification accuracy is also an important check point. So far, the classification accuracy inspection is mainly based on human-computer interaction or manual inspection in the project, which are time consuming and laborious. By harnessing the automatic high-resolution remote sensing image change detection technology based on the ERDAS IMAGINE platform, this paper carried out the classification accuracy inspection test of land cover data in the project, and presented a corresponding technical route, which includes data pre-processing, change detection, result output and information extraction. The result of the quality inspection test shows the effectiveness of the technical route, which can meet the inspection needs for the two typical errors, that is, missing and incorrect update error, and effectively reduces the work intensity of human-computer interaction inspection for quality inspectors, and also provides a technical reference for the data production and quality control of the land cover data.

Big Data: A Parallel Particle Swarm Optimization-Back-Propagation Neural Network Algorithm Based on MapReduce

PubMed Central

Cao, Jianfang; Cui, Hongyan; Shi, Hao; Jiao, Lijuan

2016-01-01

A back-propagation (BP) neural network can solve complicated random nonlinear mapping problems; therefore, it can be applied to a wide range of problems. However, as the sample size increases, the time required to train BP neural networks becomes lengthy. Moreover, the classification accuracy decreases as well. To improve the classification accuracy and runtime efficiency of the BP neural network algorithm, we proposed a parallel design and realization method for a particle swarm optimization (PSO)-optimized BP neural network based on MapReduce on the Hadoop platform using both the PSO algorithm and a parallel design. The PSO algorithm was used to optimize the BP neural network’s initial weights and thresholds and improve the accuracy of the classification algorithm. The MapReduce parallel programming model was utilized to achieve parallel processing of the BP algorithm, thereby solving the problems of hardware and communication overhead when the BP neural network addresses big data. Datasets on 5 different scales were constructed using the scene image library from the SUN Database. The classification accuracy of the parallel PSO-BP neural network algorithm is approximately 92%, and the system efficiency is approximately 0.85, which presents obvious advantages when processing big data. The algorithm proposed in this study demonstrated both higher classification accuracy and improved time efficiency, which represents a significant improvement obtained from applying parallel processing to an intelligent algorithm on big data. PMID:27304987

Hierarchical vs non-hierarchical audio indexation and classification for video genres

NASA Astrophysics Data System (ADS)

Dammak, Nouha; BenAyed, Yassine

2018-04-01

In this paper, Support Vector Machines (SVMs) are used for segmenting and indexing video genres based on only audio features extracted at block level, which has a prominent asset by capturing local temporal information. The main contribution of our study is to show the wide effect on the classification accuracies while using an hierarchical categorization structure based on Mel Frequency Cepstral Coefficients (MFCC) audio descriptor. In fact, the classification consists in three common video genres: sports videos, music clips and news scenes. The sub-classification may divide each genre into several multi-speaker and multi-dialect sub-genres. The validation of this approach was carried out on over 360 minutes of video span yielding a classification accuracy of over 99%.

Application of time series discretization using evolutionary programming for classification of precancerous cervical lesions.

PubMed

Acosta-Mesa, Héctor-Gabriel; Rechy-Ramírez, Fernando; Mezura-Montes, Efrén; Cruz-Ramírez, Nicandro; Hernández Jiménez, Rodolfo

2014-06-01

In this work, we present a novel application of time series discretization using evolutionary programming for the classification of precancerous cervical lesions. The approach optimizes the number of intervals in which the length and amplitude of the time series should be compressed, preserving the important information for classification purposes. Using evolutionary programming, the search for a good discretization scheme is guided by a cost function which considers three criteria: the entropy regarding the classification, the complexity measured as the number of different strings needed to represent the complete data set, and the compression rate assessed as the length of the discrete representation. This discretization approach is evaluated using a time series data based on temporal patterns observed during a classical test used in cervical cancer detection; the classification accuracy reached by our method is compared with the well-known times series discretization algorithm SAX and the dimensionality reduction method PCA. Statistical analysis of the classification accuracy shows that the discrete representation is as efficient as the complete raw representation for the present application, reducing the dimensionality of the time series length by 97%. This representation is also very competitive in terms of classification accuracy when compared with similar approaches. Copyright © 2014 Elsevier Inc. All rights reserved.

Cognitive-motivational deficits in ADHD: development of a classification system.

PubMed

Gupta, Rashmi; Kar, Bhoomika R; Srinivasan, Narayanan

2011-01-01

The classification systems developed so far to detect attention deficit/hyperactivity disorder (ADHD) do not have high sensitivity and specificity. We have developed a classification system based on several neuropsychological tests that measure cognitive-motivational functions that are specifically impaired in ADHD children. A total of 240 (120 ADHD children and 120 healthy controls) children in the age range of 6-9 years and 32 Oppositional Defiant Disorder (ODD) children (aged 9 years) participated in the study. Stop-Signal, Task-Switching, Attentional Network, and Choice Delay tests were administered to all the participants. Receiver operating characteristic (ROC) analysis indicated that percentage choice of long-delay reward best classified the ADHD children from healthy controls. Single parameters were not helpful in making a differential classification of ADHD with ODD. Multinominal logistic regression (MLR) was performed with multiple parameters (data fusion) that produced improved overall classification accuracy. A combination of stop-signal reaction time, posterror-slowing, mean delay, switch cost, and percentage choice of long-delay reward produced an overall classification accuracy of 97.8%; with internal validation, the overall accuracy was 92.2%. Combining parameters from different tests of control functions not only enabled us to accurately classify ADHD children from healthy controls but also in making a differential classification with ODD. These results have implications for the theories of ADHD.

Classification of EEG Signals Based on Pattern Recognition Approach.

PubMed

Amin, Hafeez Ullah; Mumtaz, Wajid; Subhani, Ahmad Rauf; Saad, Mohamad Naufal Mohamad; Malik, Aamir Saeed

2017-01-01

Feature extraction is an important step in the process of electroencephalogram (EEG) signal classification. The authors propose a "pattern recognition" approach that discriminates EEG signals recorded during different cognitive conditions. Wavelet based feature extraction such as, multi-resolution decompositions into detailed and approximate coefficients as well as relative wavelet energy were computed. Extracted relative wavelet energy features were normalized to zero mean and unit variance and then optimized using Fisher's discriminant ratio (FDR) and principal component analysis (PCA). A high density EEG dataset validated the proposed method (128-channels) by identifying two classifications: (1) EEG signals recorded during complex cognitive tasks using Raven's Advance Progressive Metric (RAPM) test; (2) EEG signals recorded during a baseline task (eyes open). Classifiers such as, K-nearest neighbors (KNN), Support Vector Machine (SVM), Multi-layer Perceptron (MLP), and Naïve Bayes (NB) were then employed. Outcomes yielded 99.11% accuracy via SVM classifier for coefficient approximations (A5) of low frequencies ranging from 0 to 3.90 Hz. Accuracy rates for detailed coefficients were 98.57 and 98.39% for SVM and KNN, respectively; and for detailed coefficients (D5) deriving from the sub-band range (3.90-7.81 Hz). Accuracy rates for MLP and NB classifiers were comparable at 97.11-89.63% and 91.60-81.07% for A5 and D5 coefficients, respectively. In addition, the proposed approach was also applied on public dataset for classification of two cognitive tasks and achieved comparable classification results, i.e., 93.33% accuracy with KNN. The proposed scheme yielded significantly higher classification performances using machine learning classifiers compared to extant quantitative feature extraction. These results suggest the proposed feature extraction method reliably classifies EEG signals recorded during cognitive tasks with a higher degree of accuracy.

Classification of EEG Signals Based on Pattern Recognition Approach

PubMed Central

Amin, Hafeez Ullah; Mumtaz, Wajid; Subhani, Ahmad Rauf; Saad, Mohamad Naufal Mohamad; Malik, Aamir Saeed

2017-01-01

Feature extraction is an important step in the process of electroencephalogram (EEG) signal classification. The authors propose a “pattern recognition” approach that discriminates EEG signals recorded during different cognitive conditions. Wavelet based feature extraction such as, multi-resolution decompositions into detailed and approximate coefficients as well as relative wavelet energy were computed. Extracted relative wavelet energy features were normalized to zero mean and unit variance and then optimized using Fisher's discriminant ratio (FDR) and principal component analysis (PCA). A high density EEG dataset validated the proposed method (128-channels) by identifying two classifications: (1) EEG signals recorded during complex cognitive tasks using Raven's Advance Progressive Metric (RAPM) test; (2) EEG signals recorded during a baseline task (eyes open). Classifiers such as, K-nearest neighbors (KNN), Support Vector Machine (SVM), Multi-layer Perceptron (MLP), and Naïve Bayes (NB) were then employed. Outcomes yielded 99.11% accuracy via SVM classifier for coefficient approximations (A5) of low frequencies ranging from 0 to 3.90 Hz. Accuracy rates for detailed coefficients were 98.57 and 98.39% for SVM and KNN, respectively; and for detailed coefficients (D5) deriving from the sub-band range (3.90–7.81 Hz). Accuracy rates for MLP and NB classifiers were comparable at 97.11–89.63% and 91.60–81.07% for A5 and D5 coefficients, respectively. In addition, the proposed approach was also applied on public dataset for classification of two cognitive tasks and achieved comparable classification results, i.e., 93.33% accuracy with KNN. The proposed scheme yielded significantly higher classification performances using machine learning classifiers compared to extant quantitative feature extraction. These results suggest the proposed feature extraction method reliably classifies EEG signals recorded during cognitive tasks with a higher degree of accuracy. PMID:29209190

AVNM: A Voting based Novel Mathematical Rule for Image Classification.

PubMed

Vidyarthi, Ankit; Mittal, Namita

2016-12-01

In machine learning, the accuracy of the system depends upon classification result. Classification accuracy plays an imperative role in various domains. Non-parametric classifier like K-Nearest Neighbor (KNN) is the most widely used classifier for pattern analysis. Besides its easiness, simplicity and effectiveness characteristics, the main problem associated with KNN classifier is the selection of a number of nearest neighbors i.e. "k" for computation. At present, it is hard to find the optimal value of "k" using any statistical algorithm, which gives perfect accuracy in terms of low misclassification error rate. Motivated by the prescribed problem, a new sample space reduction weighted voting mathematical rule (AVNM) is proposed for classification in machine learning. The proposed AVNM rule is also non-parametric in nature like KNN. AVNM uses the weighted voting mechanism with sample space reduction to learn and examine the predicted class label for unidentified sample. AVNM is free from any initial selection of predefined variable and neighbor selection as found in KNN algorithm. The proposed classifier also reduces the effect of outliers. To verify the performance of the proposed AVNM classifier, experiments are made on 10 standard datasets taken from UCI database and one manually created dataset. The experimental result shows that the proposed AVNM rule outperforms the KNN classifier and its variants. Experimentation results based on confusion matrix accuracy parameter proves higher accuracy value with AVNM rule. The proposed AVNM rule is based on sample space reduction mechanism for identification of an optimal number of nearest neighbor selections. AVNM results in better classification accuracy and minimum error rate as compared with the state-of-art algorithm, KNN, and its variants. The proposed rule automates the selection of nearest neighbor selection and improves classification rate for UCI dataset and manually created dataset. Copyright © 2016 Elsevier Ireland Ltd. All rights reserved.

Can Statistical Machine Learning Algorithms Help for Classification of Obstructive Sleep Apnea Severity to Optimal Utilization of Polysomnography Resources?

PubMed

Bozkurt, Selen; Bostanci, Asli; Turhan, Murat

2017-08-11

The goal of this study is to evaluate the results of machine learning methods for the classification of OSA severity of patients with suspected sleep disorder breathing as normal, mild, moderate and severe based on non-polysomnographic variables: 1) clinical data, 2) symptoms and 3) physical examination. In order to produce classification models for OSA severity, five different machine learning methods (Bayesian network, Decision Tree, Random Forest, Neural Networks and Logistic Regression) were trained while relevant variables and their relationships were derived empirically from observed data. Each model was trained and evaluated using 10-fold cross-validation and to evaluate classification performances of all methods, true positive rate (TPR), false positive rate (FPR), Positive Predictive Value (PPV), F measure and Area Under Receiver Operating Characteristics curve (ROC-AUC) were used. Results of 10-fold cross validated tests with different variable settings promisingly indicated that the OSA severity of suspected OSA patients can be classified, using non-polysomnographic features, with 0.71 true positive rate as the highest and, 0.15 false positive rate as the lowest, respectively. Moreover, the test results of different variables settings revealed that the accuracy of the classification models was significantly improved when physical examination variables were added to the model. Study results showed that machine learning methods can be used to estimate the probabilities of no, mild, moderate, and severe obstructive sleep apnea and such approaches may improve accurate initial OSA screening and help referring only the suspected moderate or severe OSA patients to sleep laboratories for the expensive tests.

Falls risk assessment outcomes and factors associated with falls for older Indigenous Australians.

PubMed

Hill, Keith D; Flicker, Leon; LoGiudice, Dina; Smith, Kate; Atkinson, David; Hyde, Zoë; Fenner, Stephen; Skeaf, Linda; Malay, Roslyn; Boyle, Eileen

2016-12-01

To describe the prevalence of falls and associated risk factors in older Indigenous Australians, and compare the accuracy of validated falls risk screening and assessment tools in this population in classifying fall status. Cross-sectional study of 289 Indigenous Australians aged ≥45 years from the Kimberley region of Western Australia who had a detailed assessment including self-reported falls in the past year (n=289), the adapted Elderly Falls Screening Tool (EFST; n=255), and the Falls Risk for Older People-Community (FROP-Com) screening tool (3 items, n=74) and FROP-Com falls assessment tool (n=74). 32% of participants had ≥1 fall in the preceding year, and 37.3% were classified high falls risk using the EFST (cut-off ≥2). In contrast, for the 74 participants assessed with the FROP-Com, only 14.9% were rated high risk, 35.8% moderate risk, and 49.3% low risk. The FROP-Com screen and assessment tools had the highest classification accuracy for identifying fallers in the preceding year (area under curve >0.85), with sensitivity/specificity highest for the FROP-Com assessment (cut-off ≥12), sensitivity=0.84 and specificity=0.73. Falls are common in older Indigenous Australians. The FROP-Com falls risk assessment tool appears useful in this population, and this research suggests changes that may improve its utility further. © 2016 Public Health Association of Australia.

Prostate cancer detection using machine learning techniques by employing combination of features extracting strategies.

PubMed

Hussain, Lal; Ahmed, Adeel; Saeed, Sharjil; Rathore, Saima; Awan, Imtiaz Ahmed; Shah, Saeed Arif; Majid, Abdul; Idris, Adnan; Awan, Anees Ahmed

2018-02-06

Prostate is a second leading causes of cancer deaths among men. Early detection of cancer can effectively reduce the rate of mortality caused by Prostate cancer. Due to high and multiresolution of MRIs from prostate cancer require a proper diagnostic systems and tools. In the past researchers developed Computer aided diagnosis (CAD) systems that help the radiologist to detect the abnormalities. In this research paper, we have employed novel Machine learning techniques such as Bayesian approach, Support vector machine (SVM) kernels: polynomial, radial base function (RBF) and Gaussian and Decision Tree for detecting prostate cancer. Moreover, different features extracting strategies are proposed to improve the detection performance. The features extracting strategies are based on texture, morphological, scale invariant feature transform (SIFT), and elliptic Fourier descriptors (EFDs) features. The performance was evaluated based on single as well as combination of features using Machine Learning Classification techniques. The Cross validation (Jack-knife k-fold) was performed and performance was evaluated in term of receiver operating curve (ROC) and specificity, sensitivity, Positive predictive value (PPV), negative predictive value (NPV), false positive rate (FPR). Based on single features extracting strategies, SVM Gaussian Kernel gives the highest accuracy of 98.34% with AUC of 0.999. While, using combination of features extracting strategies, SVM Gaussian kernel with texture + morphological, and EFDs + morphological features give the highest accuracy of 99.71% and AUC of 1.00.

Use of collateral information to improve LANDSAT classification accuracies

NASA Technical Reports Server (NTRS)

Strahler, A. H. (Principal Investigator)

1981-01-01

Methods to improve LANDSAT classification accuracies were investigated including: (1) the use of prior probabilities in maximum likelihood classification as a methodology to integrate discrete collateral data with continuously measured image density variables; (2) the use of the logit classifier as an alternative to multivariate normal classification that permits mixing both continuous and categorical variables in a single model and fits empirical distributions of observations more closely than the multivariate normal density function; and (3) the use of collateral data in a geographic information system as exercised to model a desired output information layer as a function of input layers of raster format collateral and image data base layers.

Classification of electroencephalograph signals using time-frequency decomposition and linear discriminant analysis

NASA Astrophysics Data System (ADS)

Szuflitowska, B.; Orlowski, P.

2017-08-01

Automated detection system consists of two key steps: extraction of features from EEG signals and classification for detection of pathology activity. The EEG sequences were analyzed using Short-Time Fourier Transform and the classification was performed using Linear Discriminant Analysis. The accuracy of the technique was tested on three sets of EEG signals: epilepsy, healthy and Alzheimer's Disease. The classification error below 10% has been considered a success. The higher accuracy are obtained for new data of unknown classes than testing data. The methodology can be helpful in differentiation epilepsy seizure and disturbances in the EEG signal in Alzheimer's Disease.

Drug related webpages classification using images and text information based on multi-kernel learning

NASA Astrophysics Data System (ADS)

Hu, Ruiguang; Xiao, Liping; Zheng, Wenjuan

2015-12-01

In this paper, multi-kernel learning(MKL) is used for drug-related webpages classification. First, body text and image-label text are extracted through HTML parsing, and valid images are chosen by the FOCARSS algorithm. Second, text based BOW model is used to generate text representation, and image-based BOW model is used to generate images representation. Last, text and images representation are fused with a few methods. Experimental results demonstrate that the classification accuracy of MKL is higher than those of all other fusion methods in decision level and feature level, and much higher than the accuracy of single-modal classification.

Landcover classification in MRF context using Dempster-Shafer fusion for multisensor imagery.

PubMed

Sarkar, Anjan; Banerjee, Anjan; Banerjee, Nilanjan; Brahma, Siddhartha; Kartikeyan, B; Chakraborty, Manab; Majumder, K L

2005-05-01

This work deals with multisensor data fusion to obtain landcover classification. The role of feature-level fusion using the Dempster-Shafer rule and that of data-level fusion in the MRF context is studied in this paper to obtain an optimally segmented image. Subsequently, segments are validated and classification accuracy for the test data is evaluated. Two examples of data fusion of optical images and a synthetic aperture radar image are presented, each set having been acquired on different dates. Classification accuracies of the technique proposed are compared with those of some recent techniques in literature for the same image data.

Random forests for classification in ecology

USGS Publications Warehouse

Cutler, D.R.; Edwards, T.C.; Beard, K.H.; Cutler, A.; Hess, K.T.; Gibson, J.; Lawler, J.J.

2007-01-01

Classification procedures are some of the most widely used statistical methods in ecology. Random forests (RF) is a new and powerful statistical classifier that is well established in other disciplines but is relatively unknown in ecology. Advantages of RF compared to other statistical classifiers include (1) very high classification accuracy; (2) a novel method of determining variable importance; (3) ability to model complex interactions among predictor variables; (4) flexibility to perform several types of statistical data analysis, including regression, classification, survival analysis, and unsupervised learning; and (5) an algorithm for imputing missing values. We compared the accuracies of RF and four other commonly used statistical classifiers using data on invasive plant species presence in Lava Beds National Monument, California, USA, rare lichen species presence in the Pacific Northwest, USA, and nest sites for cavity nesting birds in the Uinta Mountains, Utah, USA. We observed high classification accuracy in all applications as measured by cross-validation and, in the case of the lichen data, by independent test data, when comparing RF to other common classification methods. We also observed that the variables that RF identified as most important for classifying invasive plant species coincided with expectations based on the literature. ?? 2007 by the Ecological Society of America.

Application of Sensor Fusion to Improve Uav Image Classification

NASA Astrophysics Data System (ADS)

Jabari, S.; Fathollahi, F.; Zhang, Y.

2017-08-01

Image classification is one of the most important tasks of remote sensing projects including the ones that are based on using UAV images. Improving the quality of UAV images directly affects the classification results and can save a huge amount of time and effort in this area. In this study, we show that sensor fusion can improve image quality which results in increasing the accuracy of image classification. Here, we tested two sensor fusion configurations by using a Panchromatic (Pan) camera along with either a colour camera or a four-band multi-spectral (MS) camera. We use the Pan camera to benefit from its higher sensitivity and the colour or MS camera to benefit from its spectral properties. The resulting images are then compared to the ones acquired by a high resolution single Bayer-pattern colour camera (here referred to as HRC). We assessed the quality of the output images by performing image classification tests. The outputs prove that the proposed sensor fusion configurations can achieve higher accuracies compared to the images of the single Bayer-pattern colour camera. Therefore, incorporating a Pan camera on-board in the UAV missions and performing image fusion can help achieving higher quality images and accordingly higher accuracy classification results.

Convolutional Neural Network for Histopathological Analysis of Osteosarcoma.

PubMed

Mishra, Rashika; Daescu, Ovidiu; Leavey, Patrick; Rakheja, Dinesh; Sengupta, Anita

2018-03-01

Pathologists often deal with high complexity and sometimes disagreement over osteosarcoma tumor classification due to cellular heterogeneity in the dataset. Segmentation and classification of histology tissue in H&E stained tumor image datasets is a challenging task because of intra-class variations, inter-class similarity, crowded context, and noisy data. In recent years, deep learning approaches have led to encouraging results in breast cancer and prostate cancer analysis. In this article, we propose convolutional neural network (CNN) as a tool to improve efficiency and accuracy of osteosarcoma tumor classification into tumor classes (viable tumor, necrosis) versus nontumor. The proposed CNN architecture contains eight learned layers: three sets of stacked two convolutional layers interspersed with max pooling layers for feature extraction and two fully connected layers with data augmentation strategies to boost performance. The use of a neural network results in higher accuracy of average 92% for the classification. We compare the proposed architecture with three existing and proven CNN architectures for image classification: AlexNet, LeNet, and VGGNet. We also provide a pipeline to calculate percentage necrosis in a given whole slide image. We conclude that the use of neural networks can assure both high accuracy and efficiency in osteosarcoma classification.

Retinal vasculature classification using novel multifractal features

NASA Astrophysics Data System (ADS)

Ding, Y.; Ward, W. O. C.; Duan, Jinming; Auer, D. P.; Gowland, Penny; Bai, L.

2015-11-01

Retinal blood vessels have been implicated in a large number of diseases including diabetic retinopathy and cardiovascular diseases, which cause damages to retinal blood vessels. The availability of retinal vessel imaging provides an excellent opportunity for monitoring and diagnosis of retinal diseases, and automatic analysis of retinal vessels will help with the processes. However, state of the art vascular analysis methods such as counting the number of branches or measuring the curvature and diameter of individual vessels are unsuitable for the microvasculature. There has been published research using fractal analysis to calculate fractal dimensions of retinal blood vessels, but so far there has been no systematic research extracting discriminant features from retinal vessels for classifications. This paper introduces new methods for feature extraction from multifractal spectra of retinal vessels for classification. Two publicly available retinal vascular image databases are used for the experiments, and the proposed methods have produced accuracies of 85.5% and 77% for classification of healthy and diabetic retinal vasculatures. Experiments show that classification with multiple fractal features produces better rates compared with methods using a single fractal dimension value. In addition to this, experiments also show that classification accuracy can be affected by the accuracy of vessel segmentation algorithms.

A Comparative Study of Landsat TM and SPOT HRG Images for Vegetation Classification in the Brazilian Amazon.

PubMed

Lu, Dengsheng; Batistella, Mateus; de Miranda, Evaristo E; Moran, Emilio

2008-01-01

Complex forest structure and abundant tree species in the moist tropical regions often cause difficulties in classifying vegetation classes with remotely sensed data. This paper explores improvement in vegetation classification accuracies through a comparative study of different image combinations based on the integration of Landsat Thematic Mapper (TM) and SPOT High Resolution Geometric (HRG) instrument data, as well as the combination of spectral signatures and textures. A maximum likelihood classifier was used to classify the different image combinations into thematic maps. This research indicated that data fusion based on HRG multispectral and panchromatic data slightly improved vegetation classification accuracies: a 3.1 to 4.6 percent increase in the kappa coefficient compared with the classification results based on original HRG or TM multispectral images. A combination of HRG spectral signatures and two textural images improved the kappa coefficient by 6.3 percent compared with pure HRG multispectral images. The textural images based on entropy or second-moment texture measures with a window size of 9 pixels × 9 pixels played an important role in improving vegetation classification accuracy. Overall, optical remote-sensing data are still insufficient for accurate vegetation classifications in the Amazon basin.

A Comparative Study of Landsat TM and SPOT HRG Images for Vegetation Classification in the Brazilian Amazon

PubMed Central

Lu, Dengsheng; Batistella, Mateus; de Miranda, Evaristo E.; Moran, Emilio

2009-01-01

Complex forest structure and abundant tree species in the moist tropical regions often cause difficulties in classifying vegetation classes with remotely sensed data. This paper explores improvement in vegetation classification accuracies through a comparative study of different image combinations based on the integration of Landsat Thematic Mapper (TM) and SPOT High Resolution Geometric (HRG) instrument data, as well as the combination of spectral signatures and textures. A maximum likelihood classifier was used to classify the different image combinations into thematic maps. This research indicated that data fusion based on HRG multispectral and panchromatic data slightly improved vegetation classification accuracies: a 3.1 to 4.6 percent increase in the kappa coefficient compared with the classification results based on original HRG or TM multispectral images. A combination of HRG spectral signatures and two textural images improved the kappa coefficient by 6.3 percent compared with pure HRG multispectral images. The textural images based on entropy or second-moment texture measures with a window size of 9 pixels × 9 pixels played an important role in improving vegetation classification accuracy. Overall, optical remote-sensing data are still insufficient for accurate vegetation classifications in the Amazon basin. PMID:19789716

Intra- and interobserver agreement in the classification and treatment of distal third clavicle fractures.

PubMed

Bishop, Julie Y; Jones, Grant L; Lewis, Brian; Pedroza, Angela

2015-04-01

In treatment of distal third clavicle fractures, the Neer classification system, based on the location of the fracture in relation to the coracoclavicular ligaments, has traditionally been used to determine fracture pattern stability. To determine the intra- and interobserver reliability in the classification of distal third clavicle fractures via standard plain radiographs and the intra- and interobserver agreement in the preferred treatment of these fractures. Cohort study (Diagnosis); Level of evidence, 3. Thirty radiographs of distal clavicle fractures were randomly selected from patients treated for distal clavicle fractures between 2006 and 2011. The radiographs were distributed to 22 shoulder/sports medicine fellowship-trained orthopaedic surgeons. Fourteen surgeons responded and took part in the study. The evaluators were asked to measure the size of the distal fragment, classify the fracture pattern as stable or unstable, assign the Neer classification, and recommend operative versus nonoperative treatment. The radiographs were reordered and redistributed 3 months later. Inter- and intrarater agreement was determined for the distal fragment size, stability of the fracture, Neer classification, and decision to operate. Single variable logistic regression was performed to determine what factors could most accurately predict the decision for surgery. Interrater agreement was fair for distal fragment size, moderate for stability, fair for Neer classification, slight for type IIB and III fractures, and moderate for treatment approach. Intrarater agreement was moderate for distal fragment size categories (κ = 0.50, P < .001) and Neer classification (κ = 0.42, P < .001) and substantial for stable fracture (κ = 0.65, P < .001) and decision to operate (κ = 0.65, P < .001). Fracture stability was the best predictor of treatment, with 89% accuracy (P < .001). Fracture stability determination and the decision to operate had the highest interobserver agreement. Fracture stability was the key determinant of treatment, rather than the Neer classification system or the size of the distal fragment. © 2015 The Author(s).

Nearest clusters based partial least squares discriminant analysis for the classification of spectral data.

PubMed

Song, Weiran; Wang, Hui; Maguire, Paul; Nibouche, Omar

2018-06-07

Partial Least Squares Discriminant Analysis (PLS-DA) is one of the most effective multivariate analysis methods for spectral data analysis, which extracts latent variables and uses them to predict responses. In particular, it is an effective method for handling high-dimensional and collinear spectral data. However, PLS-DA does not explicitly address data multimodality, i.e., within-class multimodal distribution of data. In this paper, we present a novel method termed nearest clusters based PLS-DA (NCPLS-DA) for addressing the multimodality and nonlinearity issues explicitly and improving the performance of PLS-DA on spectral data classification. The new method applies hierarchical clustering to divide samples into clusters and calculates the corresponding centre of every cluster. For a given query point, only clusters whose centres are nearest to such a query point are used for PLS-DA. Such a method can provide a simple and effective tool for separating multimodal and nonlinear classes into clusters which are locally linear and unimodal. Experimental results on 17 datasets, including 12 UCI and 5 spectral datasets, show that NCPLS-DA can outperform 4 baseline methods, namely, PLS-DA, kernel PLS-DA, local PLS-DA and k-NN, achieving the highest classification accuracy most of the time. Copyright © 2018 Elsevier B.V. All rights reserved.

A Vegetation Database for the Colorado River Ecosystem from Glen Canyon Dam to the Western Boundary of Grand Canyon National Park, Arizona

USGS Publications Warehouse

Ralston, Barbara E.; Davis, Philip A.; Weber, Robert M.; Rundall, Jill M.

2008-01-01

A vegetation database of the riparian vegetation located within the Colorado River ecosystem (CRE), a subsection of the Colorado River between Glen Canyon Dam and the western boundary of Grand Canyon National Park, was constructed using four-band image mosaics acquired in May 2002. A digital line scanner was flown over the Colorado River corridor in Arizona by ISTAR Americas, using a Leica ADS-40 digital camera to acquire a digital surface model and four-band image mosaics (blue, green, red, and near-infrared) for vegetation mapping. The primary objective of this mapping project was to develop a digital inventory map of vegetation to enable patch- and landscape-scale change detection, and to establish randomized sampling points for ground surveys of terrestrial fauna (principally, but not exclusively, birds). The vegetation base map was constructed through a combination of ground surveys to identify vegetation classes, image processing, and automated supervised classification procedures. Analysis of the imagery and subsequent supervised classification involved multiple steps to evaluate band quality, band ratios, and vegetation texture and density. Identification of vegetation classes involved collection of cover data throughout the river corridor and subsequent analysis using two-way indicator species analysis (TWINSPAN). Vegetation was classified into six vegetation classes, following the National Vegetation Classification Standard, based on cover dominance. This analysis indicated that total area covered by all vegetation within the CRE was 3,346 ha. Considering the six vegetation classes, the sparse shrub (SS) class accounted for the greatest amount of vegetation (627 ha) followed by Pluchea (PLSE) and Tamarix (TARA) at 494 and 366 ha, respectively. The wetland (WTLD) and Prosopis-Acacia (PRGL) classes both had similar areal cover values (227 and 213 ha, respectively). Baccharis-Salix (BAXX) was the least represented at 94 ha. Accuracy assessment of the supervised classification determined that accuracies varied among vegetation classes from 90% to 49%. Causes for low accuracies were similar spectral signatures among vegetation classes. Fuzzy accuracy assessment improved classification accuracies such that Federal mapping standards of 80% accuracies for all classes were met. The scale used to quantify vegetation adequately meets the needs of the stakeholder group. Increasing the scale to meet the U.S. Geological Survey (USGS)-National Park Service (NPS)National Mapping Program's minimum mapping unit of 0.5 ha is unwarranted because this scale would reduce the resolution of some classes (e.g., seep willow/coyote willow would likely be combined with tamarisk). While this would undoubtedly improve classification accuracies, it would not provide the community-level information about vegetation change that would benefit stakeholders. The identification of vegetation classes should follow NPS mapping approaches to complement the national effort and should incorporate the alternative analysis for community identification that is being incorporated into newer NPS mapping efforts. National Vegetation Classification is followed in this report for association- to formation-level categories. Accuracies could be improved by including more environmental variables such as stage elevation in the classification process and incorporating object-based classification methods. Another approach that may address the heterogeneous species issue and classification is to use spectral mixing analysis to estimate the fractional cover of species within each pixel and better quantify the cover of individual species that compose a cover class. Varying flights to capture vegetation at different times of the year might also help separate some vegetation classes, though the cost may be prohibitive. Lastly, photointerpretation instead of automated mapping could be tried. Photointerpretation would likely not improve accuracies in this case, howev

Spatial modeling and classification of corneal shape.

PubMed

Marsolo, Keith; Twa, Michael; Bullimore, Mark A; Parthasarathy, Srinivasan

2007-03-01

One of the most promising applications of data mining is in biomedical data used in patient diagnosis. Any method of data analysis intended to support the clinical decision-making process should meet several criteria: it should capture clinically relevant features, be computationally feasible, and provide easily interpretable results. In an initial study, we examined the feasibility of using Zernike polynomials to represent biomedical instrument data in conjunction with a decision tree classifier to distinguish between the diseased and non-diseased eyes. Here, we provide a comprehensive follow-up to that work, examining a second representation, pseudo-Zernike polynomials, to determine whether they provide any increase in classification accuracy. We compare the fidelity of both methods using residual root-mean-square (rms) error and evaluate accuracy using several classifiers: neural networks, C4.5 decision trees, Voting Feature Intervals, and Naïve Bayes. We also examine the effect of several meta-learning strategies: boosting, bagging, and Random Forests (RFs). We present results comparing accuracy as it relates to dataset and transformation resolution over a larger, more challenging, multi-class dataset. They show that classification accuracy is similar for both data transformations, but differs by classifier. We find that the Zernike polynomials provide better feature representation than the pseudo-Zernikes and that the decision trees yield the best balance of classification accuracy and interpretability.

Activity classification using the GENEA: optimum sampling frequency and number of axes.

PubMed

Zhang, Shaoyan; Murray, Peter; Zillmer, Ruediger; Eston, Roger G; Catt, Michael; Rowlands, Alex V

2012-11-01

The GENEA shows high accuracy for classification of sedentary, household, walking, and running activities when sampling at 80 Hz on three axes. It is not known whether it is possible to decrease this sampling frequency and/or the number of axes without detriment to classification accuracy. The purpose of this study was to compare the classification rate of activities on the basis of data from a single axis, two axes, and three axes, with sampling rates ranging from 5 to 80 Hz. Sixty participants (age, 49.4 yr (6.5 yr); BMI, 24.6 kg·m (3.4 kg·m)) completed 10-12 semistructured activities in the laboratory and outdoor environment while wearing a GENEA accelerometer on the right wrist. We analyzed data from single axis, dual axes, and three axes at sampling rates of 5, 10, 20, 40, and 80 Hz. Mathematical models based on features extracted from mean, SD, fast Fourier transform, and wavelet decomposition were built, which combined one of the numbers of axes with one of the sampling rates to classify activities into sedentary, household, walking, and running. Classification accuracy was high irrespective of the number of axes for data collected at 80 Hz (96.93% ± 0.97%), 40 Hz (97.4% ± 0.73%), 20 Hz (96.86% ± 1.12%), and 10 Hz (97.01% ± 1.01%) but dropped for data collected at 5 Hz (94.98% ± 1.36%). Sampling frequencies >10 Hz and/or more than one axis of measurement were not associated with greater classification accuracy. Lower sampling rates and measurement of a single axis would result in a lower data load, longer battery life, and higher efficiency of data processing. Further research should investigate whether a lower sampling rate and a single axis affects classification accuracy when considering a wider range of activities.

Optimizing Support Vector Machine Parameters with Genetic Algorithm for Credit Risk Assessment

NASA Astrophysics Data System (ADS)

Manurung, Jonson; Mawengkang, Herman; Zamzami, Elviawaty

2017-12-01

Support vector machine (SVM) is a popular classification method known to have strong generalization capabilities. SVM can solve the problem of classification and linear regression or nonlinear kernel which can be a learning algorithm for the ability of classification and regression. However, SVM also has a weakness that is difficult to determine the optimal parameter value. SVM calculates the best linear separator on the input feature space according to the training data. To classify data which are non-linearly separable, SVM uses kernel tricks to transform the data into a linearly separable data on a higher dimension feature space. The kernel trick using various kinds of kernel functions, such as : linear kernel, polynomial, radial base function (RBF) and sigmoid. Each function has parameters which affect the accuracy of SVM classification. To solve the problem genetic algorithms are proposed to be applied as the optimal parameter value search algorithm thus increasing the best classification accuracy on SVM. Data taken from UCI repository of machine learning database: Australian Credit Approval. The results show that the combination of SVM and genetic algorithms is effective in improving classification accuracy. Genetic algorithms has been shown to be effective in systematically finding optimal kernel parameters for SVM, instead of randomly selected kernel parameters. The best accuracy for data has been upgraded from kernel Linear: 85.12%, polynomial: 81.76%, RBF: 77.22% Sigmoid: 78.70%. However, for bigger data sizes, this method is not practical because it takes a lot of time.

Accurate crop classification using hierarchical genetic fuzzy rule-based systems

NASA Astrophysics Data System (ADS)

Topaloglou, Charalampos A.; Mylonas, Stelios K.; Stavrakoudis, Dimitris G.; Mastorocostas, Paris A.; Theocharis, John B.

2014-10-01

This paper investigates the effectiveness of an advanced classification system for accurate crop classification using very high resolution (VHR) satellite imagery. Specifically, a recently proposed genetic fuzzy rule-based classification system (GFRBCS) is employed, namely, the Hierarchical Rule-based Linguistic Classifier (HiRLiC). HiRLiC's model comprises a small set of simple IF-THEN fuzzy rules, easily interpretable by humans. One of its most important attributes is that its learning algorithm requires minimum user interaction, since the most important learning parameters affecting the classification accuracy are determined by the learning algorithm automatically. HiRLiC is applied in a challenging crop classification task, using a SPOT5 satellite image over an intensively cultivated area in a lake-wetland ecosystem in northern Greece. A rich set of higher-order spectral and textural features is derived from the initial bands of the (pan-sharpened) image, resulting in an input space comprising 119 features. The experimental analysis proves that HiRLiC compares favorably to other interpretable classifiers of the literature, both in terms of structural complexity and classification accuracy. Its testing accuracy was very close to that obtained by complex state-of-the-art classification systems, such as the support vector machines (SVM) and random forest (RF) classifiers. Nevertheless, visual inspection of the derived classification maps shows that HiRLiC is characterized by higher generalization properties, providing more homogeneous classifications that the competitors. Moreover, the runtime requirements for producing the thematic map was orders of magnitude lower than the respective for the competitors.

Improving Pattern Recognition and Neural Network Algorithms with Applications to Solar Panel Energy Optimization

NASA Astrophysics Data System (ADS)

Zamora Ramos, Ernesto

Artificial Intelligence is a big part of automation and with today's technological advances, artificial intelligence has taken great strides towards positioning itself as the technology of the future to control, enhance and perfect automation. Computer vision includes pattern recognition and classification and machine learning. Computer vision is at the core of decision making and it is a vast and fruitful branch of artificial intelligence. In this work, we expose novel algorithms and techniques built upon existing technologies to improve pattern recognition and neural network training, initially motivated by a multidisciplinary effort to build a robot that helps maintain and optimize solar panel energy production. Our contributions detail an improved non-linear pre-processing technique to enhance poorly illuminated images based on modifications to the standard histogram equalization for an image. While the original motivation was to improve nocturnal navigation, the results have applications in surveillance, search and rescue, medical imaging enhancing, and many others. We created a vision system for precise camera distance positioning motivated to correctly locate the robot for capture of solar panel images for classification. The classification algorithm marks solar panels as clean or dirty for later processing. Our algorithm extends past image classification and, based on historical and experimental data, it identifies the optimal moment in which to perform maintenance on marked solar panels as to minimize the energy and profit loss. In order to improve upon the classification algorithm, we delved into feedforward neural networks because of their recent advancements, proven universal approximation and classification capabilities, and excellent recognition rates. We explore state-of-the-art neural network training techniques offering pointers and insights, culminating on the implementation of a complete library with support for modern deep learning architectures, multilayer percepterons and convolutional neural networks. Our research with neural networks has encountered a great deal of difficulties regarding hyperparameter estimation for good training convergence rate and accuracy. Most hyperparameters, including architecture, learning rate, regularization, trainable parameters (or weights) initialization, and so on, are chosen via a trial and error process with some educated guesses. However, we developed the first quantitative method to compare weight initialization strategies, a critical hyperparameter choice during training, to estimate among a group of candidate strategies which would make the network converge to the highest classification accuracy faster with high probability. Our method provides a quick, objective measure to compare initialization strategies to select the best possible among them beforehand without having to complete multiple training sessions for each candidate strategy to compare final results.

Geospatial Method for Computing Supplemental Multi-Decadal U.S. Coastal Land-Use and Land-Cover Classification Products, Using Landsat Data and C-CAP Products

NASA Technical Reports Server (NTRS)

Spruce, J. P.; Smoot, James; Ellis, Jean; Hilbert, Kent; Swann, Roberta

2012-01-01

This paper discusses the development and implementation of a geospatial data processing method and multi-decadal Landsat time series for computing general coastal U.S. land-use and land-cover (LULC) classifications and change products consisting of seven classes (water, barren, upland herbaceous, non-woody wetland, woody upland, woody wetland, and urban). Use of this approach extends the observational period of the NOAA-generated Coastal Change and Analysis Program (C-CAP) products by almost two decades, assuming the availability of one cloud free Landsat scene from any season for each targeted year. The Mobile Bay region in Alabama was used as a study area to develop, demonstrate, and validate the method that was applied to derive LULC products for nine dates at approximate five year intervals across a 34-year time span, using single dates of data for each classification in which forests were either leaf-on, leaf-off, or mixed senescent conditions. Classifications were computed and refined using decision rules in conjunction with unsupervised classification of Landsat data and C-CAP value-added products. Each classification's overall accuracy was assessed by comparing stratified random locations to available reference data, including higher spatial resolution satellite and aerial imagery, field survey data, and raw Landsat RGBs. Overall classification accuracies ranged from 83 to 91% with overall Kappa statistics ranging from 0.78 to 0.89. The accuracies are comparable to those from similar, generalized LULC products derived from C-CAP data. The Landsat MSS-based LULC product accuracies are similar to those from Landsat TM or ETM+ data. Accurate classifications were computed for all nine dates, yielding effective results regardless of season. This classification method yielded products that were used to compute LULC change products via additive GIS overlay techniques.

Proposal for a New Predictive Model of Short-Term Mortality After Living Donor Liver Transplantation due to Acute Liver Failure.

PubMed

Chung, Hyun Sik; Lee, Yu Jung; Jo, Yun Sung

2017-02-21

BACKGROUND Acute liver failure (ALF) is known to be a rapidly progressive and fatal disease. Various models which could help to estimate the post-transplant outcome for ALF have been developed; however, none of them have been proved to be the definitive predictive model of accuracy. We suggest a new predictive model, and investigated which model has the highest predictive accuracy for the short-term outcome in patients who underwent living donor liver transplantation (LDLT) due to ALF. MATERIAL AND METHODS Data from a total 88 patients were collected retrospectively. King's College Hospital criteria (KCH), Child-Turcotte-Pugh (CTP) classification, and model for end-stage liver disease (MELD) score were calculated. Univariate analysis was performed, and then multivariate statistical adjustment for preoperative variables of ALF prognosis was performed. A new predictive model was developed, called the MELD conjugated serum phosphorus model (MELD-p). The individual diagnostic accuracy and cut-off value of models in predicting 3-month post-transplant mortality were evaluated using the area under the receiver operating characteristic curve (AUC). The difference in AUC between MELD-p and the other models was analyzed. The diagnostic improvement in MELD-p was assessed using the net reclassification improvement (NRI) and integrated discrimination improvement (IDI). RESULTS The MELD-p and MELD scores had high predictive accuracy (AUC >0.9). KCH and serum phosphorus had an acceptable predictive ability (AUC >0.7). The CTP classification failed to show discriminative accuracy in predicting 3-month post-transplant mortality. The difference in AUC between MELD-p and the other models had statistically significant associations with CTP and KCH. The cut-off value of MELD-p was 3.98 for predicting 3-month post-transplant mortality. The NRI was 9.9% and the IDI was 2.9%. CONCLUSIONS MELD-p score can predict 3-month post-transplant mortality better than other scoring systems after LDLT due to ALF. The recommended cut-off value of MELD-p is 3.98.

Determining the best phenological state for accurate mapping of Phragmites australis in wetlands using time series multispectral satellite data

NASA Astrophysics Data System (ADS)

Rupasinghe, P. A.; Markle, C. E.; Marcaccio, J. V.; Chow-Fraser, P.

2017-12-01

Phragmites australis (European common reed), is a relatively recent invader of wetlands and beaches in Ontario. It can establish large homogenous stands within wetlands and disperse widely throughout the landscape by wind and vehicular traffic. A first step in managing this invasive species includes accurate mapping and quantification of its distribution. This is challenging because Phragimtes is distributed in a large spatial extent, which makes the mapping more costly and time consuming. Here, we used freely available multispectral satellite images taken monthly (cloud free images as available) for the calendar year to determine the optimum phenological state of Phragmites that would allow it to be accurately identified using remote sensing data. We analyzed time series, Landsat-8 OLI and Sentinel-2 images for Big Creek Wildlife Area, ON using image classification (Support Vector Machines), Normalized Difference Vegetation Index (NDVI) and Normalized Difference Water Index (NDWI). We used field sampling data and high resolution image collected using Unmanned Aerial Vehicle (UAV; 8 cm spatial resolution) as training data and for the validation of the classified images. The accuracy for all land cover classes and for Phragmites alone were low at both the start and end of the calendar year, but reached overall accuracy >85% by mid to late summer. The highest classification accuracies for Landsat-8 OLI were associated with late July and early August imagery. We observed similar trends using the Sentinel-2 images, with higher overall accuracy for all land cover classes and for Phragmites alone from late July to late September. During this period, we found the greatest difference between Phragmites and Typha, commonly confused classes, with respect to near-infrared and shortwave infrared reflectance. Therefore, the unique spectral signature of Phragmites can be attributed to both the level of greenness and factors related to water content in the leaves during late summer. Landsat-8 OLI or Sentinel-2 images acquired in late summer can be used as a cost effective approach to mapping Phragmites at a large spatial scale without sacrificing accuracy.

Classification Accuracy of Brief Parent Report Measures of Language Development in Spanish-Speaking Toddlers

ERIC Educational Resources Information Center

Guiberson, Mark; Rodriguez, Barbara L.; Dale, Philip S.

2011-01-01

Purpose: The purpose of the current study was to examine the concurrent validity and classification accuracy of 3 parent report measures of language development in Spanish-speaking toddlers. Method: Forty-five Spanish-speaking parents and their 2-year-old children participated. Twenty-three children had expressive language delays (ELDs) as…

Measurement Properties and Classification Accuracy of Two Spanish Parent Surveys of Language Development for Preschool-Age Children

ERIC Educational Resources Information Center

Guiberson, Mark; Rodriguez, Barbara L.

2010-01-01

Purpose: To describe the concurrent validity and classification accuracy of 2 Spanish parent surveys of language development, the Spanish Ages and Stages Questionnaire (ASQ; Squires, Potter, & Bricker, 1999) and the Pilot Inventario-III (Pilot INV-III; Guiberson, 2008a). Method: Forty-eight Spanish-speaking parents of preschool-age children…

A Concurrent Test of Accuracy-of-Classification for the Strong Vocational Interest and Kuder Occupational Interest Survey

ERIC Educational Resources Information Center

Zytowski, Donald G.

1972-01-01

Owing to the uncertainty concerning the concurrent validity of the SVIB and the KOIS, a test of accuracy of classification of men in the occupations common to both inventories was undertaken. The results suggest that neither show any less validity than had been shown in separate studies previously. (Author)

Using the PDD Behavior Inventory as a Level 2 Screener: A Classification and Regression Trees Analysis

ERIC Educational Resources Information Center

Cohen, Ira L.; Liu, Xudong; Hudson, Melissa; Gillis, Jennifer; Cavalari, Rachel N. S.; Romanczyk, Raymond G.; Karmel, Bernard Z.; Gardner, Judith M.

2016-01-01

In order to improve discrimination accuracy between Autism Spectrum Disorder (ASD) and similar neurodevelopmental disorders, a data mining procedure, Classification and Regression Trees (CART), was used on a large multi-site sample of PDD Behavior Inventory (PDDBI) forms on children with and without ASD. Discrimination accuracy exceeded 80%,…

Developing Local Oral Reading Fluency Cut Scores for Predicting High-Stakes Test Performance

ERIC Educational Resources Information Center

Grapin, Sally L.; Kranzler, John H.; Waldron, Nancy; Joyce-Beaulieu, Diana; Algina, James

2017-01-01

This study evaluated the classification accuracy of a second grade oral reading fluency curriculum-based measure (R-CBM) in predicting third grade state test performance. It also compared the long-term classification accuracy of local and publisher-recommended R-CBM cut scores. Participants were 266 students who were divided into a calibration…

Factors Affecting the Item Parameter Estimation and Classification Accuracy of the DINA Model

ERIC Educational Resources Information Center

de la Torre, Jimmy; Hong, Yuan; Deng, Weiling

2010-01-01

To better understand the statistical properties of the deterministic inputs, noisy "and" gate cognitive diagnosis (DINA) model, the impact of several factors on the quality of the item parameter estimates and classification accuracy was investigated. Results of the simulation study indicate that the fully Bayes approach is most accurate when the…

Predictive modeling of respiratory tumor motion for real-time prediction of baseline shifts

NASA Astrophysics Data System (ADS)

Balasubramanian, A.; Shamsuddin, R.; Prabhakaran, B.; Sawant, A.

2017-03-01

Baseline shifts in respiratory patterns can result in significant spatiotemporal changes in patient anatomy (compared to that captured during simulation), in turn, causing geometric and dosimetric errors in the administration of thoracic and abdominal radiotherapy. We propose predictive modeling of the tumor motion trajectories for predicting a baseline shift ahead of its occurrence. The key idea is to use the features of the tumor motion trajectory over a 1 min window, and predict the occurrence of a baseline shift in the 5 s that immediately follow (lookahead window). In this study, we explored a preliminary trend-based analysis with multi-class annotations as well as a more focused binary classification analysis. In both analyses, a number of different inter-fraction and intra-fraction training strategies were studied, both offline as well as online, along with data sufficiency and skew compensation for class imbalances. The performance of different training strategies were compared across multiple machine learning classification algorithms, including nearest neighbor, Naïve Bayes, linear discriminant and ensemble Adaboost. The prediction performance is evaluated using metrics such as accuracy, precision, recall and the area under the curve (AUC) for repeater operating characteristics curve. The key results of the trend-based analysis indicate that (i) intra-fraction training strategies achieve highest prediction accuracies (90.5-91.4%) (ii) the predictive modeling yields lowest accuracies (50-60%) when the training data does not include any information from the test patient; (iii) the prediction latencies are as low as a few hundred milliseconds, and thus conducive for real-time prediction. The binary classification performance is promising, indicated by high AUCs (0.96-0.98). It also confirms the utility of prior data from previous patients, and also the necessity of training the classifier on some initial data from the new patient for reasonable prediction performance. The ability to predict a baseline shift with a sufficient look-ahead window will enable clinical systems or even human users to hold the treatment beam in such situations, thereby reducing the probability of serious geometric and dosimetric errors.

EEG complexity as a biomarker for autism spectrum disorder risk

PubMed Central

2011-01-01

Background Complex neurodevelopmental disorders may be characterized by subtle brain function signatures early in life before behavioral symptoms are apparent. Such endophenotypes may be measurable biomarkers for later cognitive impairments. The nonlinear complexity of electroencephalography (EEG) signals is believed to contain information about the architecture of the neural networks in the brain on many scales. Early detection of abnormalities in EEG signals may be an early biomarker for developmental cognitive disorders. The goal of this paper is to demonstrate that the modified multiscale entropy (mMSE) computed on the basis of resting state EEG data can be used as a biomarker of normal brain development and distinguish typically developing children from a group of infants at high risk for autism spectrum disorder (ASD), defined on the basis of an older sibling with ASD. Methods Using mMSE as a feature vector, a multiclass support vector machine algorithm was used to classify typically developing and high-risk groups. Classification was computed separately within each age group from 6 to 24 months. Results Multiscale entropy appears to go through a different developmental trajectory in infants at high risk for autism (HRA) than it does in typically developing controls. Differences appear to be greatest at ages 9 to 12 months. Using several machine learning algorithms with mMSE as a feature vector, infants were classified with over 80% accuracy into control and HRA groups at age 9 months. Classification accuracy for boys was close to 100% at age 9 months and remains high (70% to 90%) at ages 12 and 18 months. For girls, classification accuracy was highest at age 6 months, but declines thereafter. Conclusions This proof-of-principle study suggests that mMSE computed from resting state EEG signals may be a useful biomarker for early detection of risk for ASD and abnormalities in cognitive development in infants. To our knowledge, this is the first demonstration of an information theoretic analysis of EEG data for biomarkers in infants at risk for a complex neurodevelopmental disorder. PMID:21342500

Predictive modeling of respiratory tumor motion for real-time prediction of baseline shifts

PubMed Central

Balasubramanian, A; Shamsuddin, R; Prabhakaran, B; Sawant, A

2017-01-01

Baseline shifts in respiratory patterns can result in significant spatiotemporal changes in patient anatomy (compared to that captured during simulation), in turn, causing geometric and dosimetric errors in the administration of thoracic and abdominal radiotherapy. We propose predictive modeling of the tumor motion trajectories for predicting a baseline shift ahead of its occurrence. The key idea is to use the features of the tumor motion trajectory over a 1 min window, and predict the occurrence of a baseline shift in the 5 s that immediately follow (lookahead window). In this study, we explored a preliminary trend-based analysis with multi-class annotations as well as a more focused binary classification analysis. In both analyses, a number of different inter-fraction and intra-fraction training strategies were studied, both offline as well as online, along with data sufficiency and skew compensation for class imbalances. The performance of different training strategies were compared across multiple machine learning classification algorithms, including nearest neighbor, Naïve Bayes, linear discriminant and ensemble Adaboost. The prediction performance is evaluated using metrics such as accuracy, precision, recall and the area under the curve (AUC) for repeater operating characteristics curve. The key results of the trend-based analysis indicate that (i) intra-fraction training strategies achieve highest prediction accuracies (90.5–91.4%); (ii) the predictive modeling yields lowest accuracies (50–60%) when the training data does not include any information from the test patient; (iii) the prediction latencies are as low as a few hundred milliseconds, and thus conducive for real-time prediction. The binary classification performance is promising, indicated by high AUCs (0.96–0.98). It also confirms the utility of prior data from previous patients, and also the necessity of training the classifier on some initial data from the new patient for reasonable prediction performance. The ability to predict a baseline shift with a sufficient lookahead window will enable clinical systems or even human users to hold the treatment beam in such situations, thereby reducing the probability of serious geometric and dosimetric errors. PMID:28075331

Predictive modeling of respiratory tumor motion for real-time prediction of baseline shifts.

PubMed

Balasubramanian, A; Shamsuddin, R; Prabhakaran, B; Sawant, A

2017-03-07

Baseline shifts in respiratory patterns can result in significant spatiotemporal changes in patient anatomy (compared to that captured during simulation), in turn, causing geometric and dosimetric errors in the administration of thoracic and abdominal radiotherapy. We propose predictive modeling of the tumor motion trajectories for predicting a baseline shift ahead of its occurrence. The key idea is to use the features of the tumor motion trajectory over a 1 min window, and predict the occurrence of a baseline shift in the 5 s that immediately follow (lookahead window). In this study, we explored a preliminary trend-based analysis with multi-class annotations as well as a more focused binary classification analysis. In both analyses, a number of different inter-fraction and intra-fraction training strategies were studied, both offline as well as online, along with data sufficiency and skew compensation for class imbalances. The performance of different training strategies were compared across multiple machine learning classification algorithms, including nearest neighbor, Naïve Bayes, linear discriminant and ensemble Adaboost. The prediction performance is evaluated using metrics such as accuracy, precision, recall and the area under the curve (AUC) for repeater operating characteristics curve. The key results of the trend-based analysis indicate that (i) intra-fraction training strategies achieve highest prediction accuracies (90.5-91.4%); (ii) the predictive modeling yields lowest accuracies (50-60%) when the training data does not include any information from the test patient; (iii) the prediction latencies are as low as a few hundred milliseconds, and thus conducive for real-time prediction. The binary classification performance is promising, indicated by high AUCs (0.96-0.98). It also confirms the utility of prior data from previous patients, and also the necessity of training the classifier on some initial data from the new patient for reasonable prediction performance. The ability to predict a baseline shift with a sufficient look-ahead window will enable clinical systems or even human users to hold the treatment beam in such situations, thereby reducing the probability of serious geometric and dosimetric errors.

Faster Trees: Strategies for Accelerated Training and Prediction of Random Forests for Classification of Polsar Images

NASA Astrophysics Data System (ADS)

Hänsch, Ronny; Hellwich, Olaf

2018-04-01

Random Forests have continuously proven to be one of the most accurate, robust, as well as efficient methods for the supervised classification of images in general and polarimetric synthetic aperture radar data in particular. While the majority of previous work focus on improving classification accuracy, we aim for accelerating the training of the classifier as well as its usage during prediction while maintaining its accuracy. Unlike other approaches we mainly consider algorithmic changes to stay as much as possible independent of platform and programming language. The final model achieves an approximately 60 times faster training and a 500 times faster prediction, while the accuracy is only marginally decreased by roughly 1 %.

How reliable and accurate is the AO/OTA comprehensive classification for adult long-bone fractures?

PubMed

Meling, Terje; Harboe, Knut; Enoksen, Cathrine H; Aarflot, Morten; Arthursson, Astvaldur J; Søreide, Kjetil

2012-07-01

Reliable classification of fractures is important for treatment allocation and study comparisons. The overall accuracy of scoring applied to a general population of fractures is little known. This study aimed to investigate the accuracy and reliability of the comprehensive Arbeitsgemeinschaft für Osteosynthesefragen/Orthopedic Trauma Association classification for adult long-bone fractures and identify factors associated with poor coding agreement. Adults (>16 years) with long-bone fractures coded in a Fracture and Dislocation Registry at the Stavanger University Hospital during the fiscal year 2008 were included. An unblinded reference code dataset was generated for the overall accuracy assessment by two experienced orthopedic trauma surgeons. Blinded analysis of intrarater reliability was performed by rescoring and of interrater reliability by recoding of a randomly selected fracture sample. Proportion of agreement (PA) and kappa (κ) statistics are presented. Uni- and multivariate logistic regression analyses of factors predicting accuracy were performed. During the study period, 949 fractures were included and coded by 26 surgeons. For the intrarater analysis, overall agreements were κ = 0.67 (95% confidence interval [CI]: 0.64-0.70) and PA 69%. For interrater assessment, κ = 0.67 (95% CI: 0.62-0.72) and PA 69%. The accuracy of surgeons' blinded recoding was κ = 0.68 (95% CI: 0.65- 0.71) and PA 68%. Fracture type, frequency of the fracture, and segment fractured significantly influenced accuracy whereas the coder's experience did not. Both the reliability and accuracy of the comprehensive Arbeitsgemeinschaft für Osteosynthesefragen/Orthopedic Trauma Association classification for long-bone fractures ranged from substantial to excellent. Variations in coding accuracy seem to be related more to the fracture itself than the surgeon. Diagnostic study, level I.

Discriminative Hierarchical K-Means Tree for Large-Scale Image Classification.

PubMed

Chen, Shizhi; Yang, Xiaodong; Tian, Yingli

2015-09-01

A key challenge in large-scale image classification is how to achieve efficiency in terms of both computation and memory without compromising classification accuracy. The learning-based classifiers achieve the state-of-the-art accuracies, but have been criticized for the computational complexity that grows linearly with the number of classes. The nonparametric nearest neighbor (NN)-based classifiers naturally handle large numbers of categories, but incur prohibitively expensive computation and memory costs. In this brief, we present a novel classification scheme, i.e., discriminative hierarchical K-means tree (D-HKTree), which combines the advantages of both learning-based and NN-based classifiers. The complexity of the D-HKTree only grows sublinearly with the number of categories, which is much better than the recent hierarchical support vector machines-based methods. The memory requirement is the order of magnitude less than the recent Naïve Bayesian NN-based approaches. The proposed D-HKTree classification scheme is evaluated on several challenging benchmark databases and achieves the state-of-the-art accuracies, while with significantly lower computation cost and memory requirement.

Polsar Land Cover Classification Based on Hidden Polarimetric Features in Rotation Domain and Svm Classifier

NASA Astrophysics Data System (ADS)

Tao, C.-S.; Chen, S.-W.; Li, Y.-Z.; Xiao, S.-P.

2017-09-01

Land cover classification is an important application for polarimetric synthetic aperture radar (PolSAR) data utilization. Rollinvariant polarimetric features such as H / Ani / α / Span are commonly adopted in PolSAR land cover classification. However, target orientation diversity effect makes PolSAR images understanding and interpretation difficult. Only using the roll-invariant polarimetric features may introduce ambiguity in the interpretation of targets' scattering mechanisms and limit the followed classification accuracy. To address this problem, this work firstly focuses on hidden polarimetric feature mining in the rotation domain along the radar line of sight using the recently reported uniform polarimetric matrix rotation theory and the visualization and characterization tool of polarimetric coherence pattern. The former rotates the acquired polarimetric matrix along the radar line of sight and fully describes the rotation characteristics of each entry of the matrix. Sets of new polarimetric features are derived to describe the hidden scattering information of the target in the rotation domain. The latter extends the traditional polarimetric coherence at a given rotation angle to the rotation domain for complete interpretation. A visualization and characterization tool is established to derive new polarimetric features for hidden information exploration. Then, a classification scheme is developed combing both the selected new hidden polarimetric features in rotation domain and the commonly used roll-invariant polarimetric features with a support vector machine (SVM) classifier. Comparison experiments based on AIRSAR and multi-temporal UAVSAR data demonstrate that compared with the conventional classification scheme which only uses the roll-invariant polarimetric features, the proposed classification scheme achieves both higher classification accuracy and better robustness. For AIRSAR data, the overall classification accuracy with the proposed classification scheme is 94.91 %, while that with the conventional classification scheme is 93.70 %. Moreover, for multi-temporal UAVSAR data, the averaged overall classification accuracy with the proposed classification scheme is up to 97.08 %, which is much higher than the 87.79 % from the conventional classification scheme. Furthermore, for multitemporal PolSAR data, the proposed classification scheme can achieve better robustness. The comparison studies also clearly demonstrate that mining and utilization of hidden polarimetric features and information in the rotation domain can gain the added benefits for PolSAR land cover classification and provide a new vision for PolSAR image interpretation and application.

Analysis of spatial distribution of land cover maps accuracy

NASA Astrophysics Data System (ADS)

Khatami, R.; Mountrakis, G.; Stehman, S. V.

2017-12-01

Land cover maps have become one of the most important products of remote sensing science. However, classification errors will exist in any classified map and affect the reliability of subsequent map usage. Moreover, classification accuracy often varies over different regions of a classified map. These variations of accuracy will affect the reliability of subsequent analyses of different regions based on the classified maps. The traditional approach of map accuracy assessment based on an error matrix does not capture the spatial variation in classification accuracy. Here, per-pixel accuracy prediction methods are proposed based on interpolating accuracy values from a test sample to produce wall-to-wall accuracy maps. Different accuracy prediction methods were developed based on four factors: predictive domain (spatial versus spectral), interpolation function (constant, linear, Gaussian, and logistic), incorporation of class information (interpolating each class separately versus grouping them together), and sample size. Incorporation of spectral domain as explanatory feature spaces of classification accuracy interpolation was done for the first time in this research. Performance of the prediction methods was evaluated using 26 test blocks, with 10 km × 10 km dimensions, dispersed throughout the United States. The performance of the predictions was evaluated using the area under the curve (AUC) of the receiver operating characteristic. Relative to existing accuracy prediction methods, our proposed methods resulted in improvements of AUC of 0.15 or greater. Evaluation of the four factors comprising the accuracy prediction methods demonstrated that: i) interpolations should be done separately for each class instead of grouping all classes together; ii) if an all-classes approach is used, the spectral domain will result in substantially greater AUC than the spatial domain; iii) for the smaller sample size and per-class predictions, the spectral and spatial domain yielded similar AUC; iv) for the larger sample size (i.e., very dense spatial sample) and per-class predictions, the spatial domain yielded larger AUC; v) increasing the sample size improved accuracy predictions with a greater benefit accruing to the spatial domain; and vi) the function used for interpolation had the smallest effect on AUC.

Liver fibrosis diagnosis by blood test and elastography in chronic hepatitis C: agreement or combination?

PubMed

Calès, P; Boursier, J; Lebigot, J; de Ledinghen, V; Aubé, C; Hubert, I; Oberti, F

2017-04-01

In chronic hepatitis C, the European Association for the Study of the Liver and the Asociacion Latinoamericana para el Estudio del Higado recommend performing transient elastography plus a blood test to diagnose significant fibrosis; test concordance confirms the diagnosis. To validate this rule and improve it by combining a blood test, FibroMeter (virus second generation, Echosens, Paris, France) and transient elastography (constitutive tests) into a single combined test, as suggested by the American Association for the Study of Liver Diseases and the Infectious Diseases Society of America. A total of 1199 patients were included in an exploratory set (HCV, n = 679) or in two validation sets (HCV ± HIV, HBV, n = 520). Accuracy was mainly evaluated by correct diagnosis rate for severe fibrosis (pathological Metavir F ≥ 3, primary outcome) by classical test scores or a fibrosis classification, reflecting Metavir staging, as a function of test concordance. Score accuracy: there were no significant differences between the blood test (75.7%), elastography (79.1%) and the combined test (79.4%) (P = 0.066); the score accuracy of each test was significantly (P < 0.001) decreased in discordant vs. concordant tests. Classification accuracy: combined test accuracy (91.7%) was significantly (P < 0.001) increased vs. the blood test (84.1%) and elastography (88.2%); accuracy of each constitutive test was significantly (P < 0.001) decreased in discordant vs. concordant tests but not with combined test: 89.0 vs. 92.7% (P = 0.118). Multivariate analysis for accuracy showed an interaction between concordance and fibrosis level: in the 1% of patients with full classification discordance and severe fibrosis, non-invasive tests were unreliable. The advantage of combined test classification was confirmed in the validation sets. The concordance recommendation is validated. A combined test, expressed in classification instead of score, improves this rule and validates the recommendation of a combined test, avoiding 99% of biopsies, and offering precise staging. © 2017 John Wiley & Sons Ltd.

Combined use of SAR and optical data for environmental assessments around refugee camps in semiarid landscapes

NASA Astrophysics Data System (ADS)

Braun, A.; Hochschild, V.

2015-04-01

Over 15 million people were officially considered as refugees in the year 2012 and another 28 million as internally displaced people (IDPs). Natural disasters, climatic and environmental changes, violent regional conflicts and population growth force people to migrate in all parts of this world. This trend is likely to continue in the near future, as political instabilities increase and land degradation progresses. EO4HumEn aims at developing operational services to support humanitarian operations during crisis situations by means of dedicated geo-spatial information products derived from Earth observation and GIS data. The goal is to develop robust, automated methods of image analysis routines for population estimation, identification of potential groundwater extraction sites and monitoring the environmental impact of refugee/IDP camps. This study investigates the combination of satellite SAR data with optical sensors and elevation information for the assessment of the environmental conditions around refugee camps. In order to estimate their impact on land degradation, land cover classifications are required which target dynamic landscapes. We performed a land use / land cover classification based on a random forest algorithm and 39 input prediction rasters based on Landsat 8 data and additional layers generated from radar texture and elevation information. The overall accuracy was 92.9 %, while optical data had the highest impact on the final classification. By analysing all combinations of the three input datasets we additionally estimated their impact on single classification outcomes and land cover classes.

Classification of Baltic Sea ice types by airborne multifrequency microwave radiometer

DOE Office of Scientific and Technical Information (OSTI.GOV)

Kurvonen, L.; Hallikainen, M.

An airborne multifrequency radiometer (24, 34, 48, and 94 GHz, vertical polarization) was used to investigate the behavior of the brightness temperature of different sea ice types in the Gulf of Bothnia (Baltic Sea). The measurements and the main results of the analysis are presented. The measurements were made in dry and wet conditions (air temperature above and below 0 C). The angle of incidence was 45{degree} in all measurements. The following topics are evaluated: (a) frequency dependency of the brightness temperature of different ice types, (b) the capability of the multifrequency radiometer to classify ice types for winter navigationmore » purposes, and (c) the optimum measurement frequencies for mapping sea ice. The weather conditions had a significant impact on the radiometric signatures of some ice types (snow-covered compact pack ice and frost-covered new ice); the impact was the highest at 94 GHz. In all cases the overall classification accuracy was around 90% (the kappa coefficient was from 0.86 to 0.96) when the optimum channel combination (24/34 GHz and 94 GHz) was used.« less

The classification of almonds (Prunus dulcis) by country and variety using UHPLC-HRMS-based untargeted metabolomics.

PubMed

Gil Solsona, R; Boix, C; Ibáñez, M; Sancho, J V

2018-03-01

The aim of this study was to use an untargeted UHPLC-HRMS-based metabolomics approach allowing discrimination between almonds based on their origin and variety. Samples were homogenised, extracted with ACN:H 2 O (80:20) containing 0.1% HCOOH and injected in a UHPLC-QTOF instrument in both positive and negative ionisation modes. Principal component analysis (PCA) was performed to ensure the absence of outliers. Partial least squares - discriminant analysis (PLS-DA) was employed to create and validate the models for country (with five different compounds) and variety (with 20 features), showing more than 95% accuracy. Additional samples were injected and the model was evaluated with blind samples, with more than 95% of samples being correctly classified using both models. MS/MS experiments were carried out to tentatively elucidate the highlighted marker compounds (pyranosides, peptides or amino acids, among others). This study has shown the potential of high-resolution mass spectrometry to perform and validate classification models, also providing information concerning the identification of the unexpected biomarkers which showed the highest discriminant power.

Development of a two-stage gene selection method that incorporates a novel hybrid approach using the cuckoo optimization algorithm and harmony search for cancer classification.

PubMed

Elyasigomari, V; Lee, D A; Screen, H R C; Shaheed, M H

2017-03-01

For each cancer type, only a few genes are informative. Due to the so-called 'curse of dimensionality' problem, the gene selection task remains a challenge. To overcome this problem, we propose a two-stage gene selection method called MRMR-COA-HS. In the first stage, the minimum redundancy and maximum relevance (MRMR) feature selection is used to select a subset of relevant genes. The selected genes are then fed into a wrapper setup that combines a new algorithm, COA-HS, using the support vector machine as a classifier. The method was applied to four microarray datasets, and the performance was assessed by the leave one out cross-validation method. Comparative performance assessment of the proposed method with other evolutionary algorithms suggested that the proposed algorithm significantly outperforms other methods in selecting a fewer number of genes while maintaining the highest classification accuracy. The functions of the selected genes were further investigated, and it was confirmed that the selected genes are biologically relevant to each cancer type. Copyright © 2017. Published by Elsevier Inc.

Simulation of seagrass bed mapping by satellite images based on the radiative transfer model

NASA Astrophysics Data System (ADS)

Sagawa, Tatsuyuki; Komatsu, Teruhisa

2015-06-01

Seagrass and seaweed beds play important roles in coastal marine ecosystems. They are food sources and habitats for many marine organisms, and influence the physical, chemical, and biological environment. They are sensitive to human impacts such as reclamation and pollution. Therefore, their management and preservation are necessary for a healthy coastal environment. Satellite remote sensing is a useful tool for mapping and monitoring seagrass beds. The efficiency of seagrass mapping, seagrass bed classification in particular, has been evaluated by mapping accuracy using an error matrix. However, mapping accuracies are influenced by coastal environments such as seawater transparency, bathymetry, and substrate type. Coastal management requires sufficient accuracy and an understanding of mapping limitations for monitoring coastal habitats including seagrass beds. Previous studies are mainly based on case studies in specific regions and seasons. Extensive data are required to generalise assessments of classification accuracy from case studies, which has proven difficult. This study aims to build a simulator based on a radiative transfer model to produce modelled satellite images and assess the visual detectability of seagrass beds under different transparencies and seagrass coverages, as well as to examine mapping limitations and classification accuracy. Our simulations led to the development of a model of water transparency and the mapping of depth limits and indicated the possibility for seagrass density mapping under certain ideal conditions. The results show that modelling satellite images is useful in evaluating the accuracy of classification and that establishing seagrass bed monitoring by remote sensing is a reliable tool.

Influence of multi-source and multi-temporal remotely sensed and ancillary data on the accuracy of random forest classification of wetlands in northern Minnesota

USGS Publications Warehouse

Corcoran, Jennifer M.; Knight, Joseph F.; Gallant, Alisa L.

2013-01-01

Wetland mapping at the landscape scale using remotely sensed data requires both affordable data and an efficient accurate classification method. Random forest classification offers several advantages over traditional land cover classification techniques, including a bootstrapping technique to generate robust estimations of outliers in the training data, as well as the capability of measuring classification confidence. Though the random forest classifier can generate complex decision trees with a multitude of input data and still not run a high risk of over fitting, there is a great need to reduce computational and operational costs by including only key input data sets without sacrificing a significant level of accuracy. Our main questions for this study site in Northern Minnesota were: (1) how does classification accuracy and confidence of mapping wetlands compare using different remote sensing platforms and sets of input data; (2) what are the key input variables for accurate differentiation of upland, water, and wetlands, including wetland type; and (3) which datasets and seasonal imagery yield the best accuracy for wetland classification. Our results show the key input variables include terrain (elevation and curvature) and soils descriptors (hydric), along with an assortment of remotely sensed data collected in the spring (satellite visible, near infrared, and thermal bands; satellite normalized vegetation index and Tasseled Cap greenness and wetness; and horizontal-horizontal (HH) and horizontal-vertical (HV) polarization using L-band satellite radar). We undertook this exploratory analysis to inform decisions by natural resource managers charged with monitoring wetland ecosystems and to aid in designing a system for consistent operational mapping of wetlands across landscapes similar to those found in Northern Minnesota.

A Visual mining based framework for classification accuracy estimation

NASA Astrophysics Data System (ADS)

Arun, Pattathal Vijayakumar

2013-12-01

Classification techniques have been widely used in different remote sensing applications and correct classification of mixed pixels is a tedious task. Traditional approaches adopt various statistical parameters, however does not facilitate effective visualisation. Data mining tools are proving very helpful in the classification process. We propose a visual mining based frame work for accuracy assessment of classification techniques using open source tools such as WEKA and PREFUSE. These tools in integration can provide an efficient approach for getting information about improvements in the classification accuracy and helps in refining training data set. We have illustrated framework for investigating the effects of various resampling methods on classification accuracy and found that bilinear (BL) is best suited for preserving radiometric characteristics. We have also investigated the optimal number of folds required for effective analysis of LISS-IV images. Techniki klasyfikacji są szeroko wykorzystywane w różnych aplikacjach teledetekcyjnych, w których poprawna klasyfikacja pikseli stanowi poważne wyzwanie. Podejście tradycyjne wykorzystujące różnego rodzaju parametry statystyczne nie zapewnia efektywnej wizualizacji. Wielce obiecujące wydaje się zastosowanie do klasyfikacji narzędzi do eksploracji danych. W artykule zaproponowano podejście bazujące na wizualnej analizie eksploracyjnej, wykorzystujące takie narzędzia typu open source jak WEKA i PREFUSE. Wymienione narzędzia ułatwiają korektę pół treningowych i efektywnie wspomagają poprawę dokładności klasyfikacji. Działanie metody sprawdzono wykorzystując wpływ różnych metod resampling na zachowanie dokładności radiometrycznej i uzyskując najlepsze wyniki dla metody bilinearnej (BL).

An initial analysis of LANDSAT 4 Thematic Mapper data for the classification of agricultural, forested wetland, and urban land covers

NASA Technical Reports Server (NTRS)

Quattrochi, D. A.; Anderson, J. E.; Brannon, D. P.; Hill, C. L.

1982-01-01

An initial analysis of LANDSAT 4 thematic mapper (TM) data for the delineation and classification of agricultural, forested wetland, and urban land covers was conducted. A study area in Poinsett County, Arkansas was used to evaluate a classification of agricultural lands derived from multitemporal LANDSAT multispectral scanner (MSS) data in comparison with a classification of TM data for the same area. Data over Reelfoot Lake in northwestern Tennessee were utilized to evaluate the TM for delineating forested wetland species. A classification of the study area was assessed for accuracy in discriminating five forested wetland categories. Finally, the TM data were used to identify urban features within a small city. A computer generated classification of Union City, Tennessee was analyzed for accuracy in delineating urban land covers. An evaluation of digitally enhanced TM data using principal components analysis to facilitate photointerpretation of urban features was also performed.

LANDSAT applications to wetlands classification in the upper Mississippi River Valley. Ph.D. Thesis. Final Report

NASA Technical Reports Server (NTRS)

Lillesand, T. M.; Werth, L. F. (Principal Investigator)

1980-01-01

A 25% improvement in average classification accuracy was realized by processing double-date vs. single-date data. Under the spectrally and spatially complex site conditions characterizing the geographical area used, further improvement in wetland classification accuracy is apparently precluded by the spectral and spatial resolution restrictions of the LANDSAT MSS. Full scene analysis of scanning densitometer data extracted from scale infrared photography failed to permit discrimination of many wetland and nonwetland cover types. When classification of photographic data was limited to wetland areas only, much more detailed and accurate classification could be made. The integration of conventional image interpretation (to simply delineate wetland boundaries) and machine assisted classification (to discriminate among cover types present within the wetland areas) appears to warrant further research to study the feasibility and cost of extending this methodology over a large area using LANDSAT and/or small scale photography.

Machine learning classifier using abnormal brain network topological metrics in major depressive disorder.

PubMed

Guo, Hao; Cao, Xiaohua; Liu, Zhifen; Li, Haifang; Chen, Junjie; Zhang, Kerang

2012-12-05

Resting state functional brain networks have been widely studied in brain disease research. However, it is currently unclear whether abnormal resting state functional brain network metrics can be used with machine learning for the classification of brain diseases. Resting state functional brain networks were constructed for 28 healthy controls and 38 major depressive disorder patients by thresholding partial correlation matrices of 90 regions. Three nodal metrics were calculated using graph theory-based approaches. Nonparametric permutation tests were then used for group comparisons of topological metrics, which were used as classified features in six different algorithms. We used statistical significance as the threshold for selecting features and measured the accuracies of six classifiers with different number of features. A sensitivity analysis method was used to evaluate the importance of different features. The result indicated that some of the regions exhibited significantly abnormal nodal centralities, including the limbic system, basal ganglia, medial temporal, and prefrontal regions. Support vector machine with radial basis kernel function algorithm and neural network algorithm exhibited the highest average accuracy (79.27 and 78.22%, respectively) with 28 features (P<0.05). Correlation analysis between feature importance and the statistical significance of metrics was investigated, and the results revealed a strong positive correlation between them. Overall, the current study demonstrated that major depressive disorder is associated with abnormal functional brain network topological metrics and statistically significant nodal metrics can be successfully used for feature selection in classification algorithms.

Online hyperspectral imaging system for evaluating quality of agricultural products

NASA Astrophysics Data System (ADS)

Mo, Changyeun; Kim, Giyoung; Lim, Jongguk

2017-06-01

The consumption of fresh-cut agricultural produce in Korea has been growing. The browning of fresh-cut vegetables that occurs during storage and foreign substances such as worms and slugs are some of the main causes of consumers' concerns with respect to safety and hygiene. The purpose of this study is to develop an on-line system for evaluating quality of agricultural products using hyperspectral imaging technology. The online evaluation system with single visible-near infrared hyperspectral camera in the range of 400 nm to 1000 nm that can assess quality of both surfaces of agricultural products such as fresh-cut lettuce was designed. Algorithms to detect browning surface were developed for this system. The optimal wavebands for discriminating between browning and sound lettuce as well as between browning lettuce and the conveyor belt were investigated using the correlation analysis and the one-way analysis of variance method. The imaging algorithms to discriminate the browning lettuces were developed using the optimal wavebands. The ratio image (RI) algorithm of the 533 nm and 697 nm images (RI533/697) for abaxial surface lettuce and the ratio image algorithm (RI533/697) and subtraction image (SI) algorithm (SI538-697) for adaxial surface lettuce had the highest classification accuracies. The classification accuracy of browning and sound lettuce was 100.0% and above 96.0%, respectively, for the both surfaces. The overall results show that the online hyperspectral imaging system could potentially be used to assess quality of agricultural products.

Characterization and delineation of caribou habitat on Unimak Island using remote sensing techniques

NASA Astrophysics Data System (ADS)

Atkinson, Brain M.

The assessment of herbivore habitat quality is traditionally based on quantifying the forages available to the animal across their home range through ground-based techniques. While these methods are highly accurate, they can be time-consuming and highly expensive, especially for herbivores that occupy vast spatial landscapes. The Unimak Island caribou herd has been decreasing in the last decade at rates that have prompted discussion of management intervention. Frequent inclement weather in this region of Alaska has provided for little opportunity to study the caribou forage habitat on Unimak Island. The overall objectives of this study were two-fold 1) to assess the feasibility of using high-resolution color and near-infrared aerial imagery to map the forage distribution of caribou habitat on Unimak Island and 2) to assess the use of a new high-resolution multispectral satellite imagery platform, RapidEye, and use of the "red-edge" spectral band on vegetation classification accuracy. Maximum likelihood classification algorithms were used to create land cover maps in aerial and satellite imagery. Accuracy assessments and transformed divergence values were produced to assess vegetative spectral information and classification accuracy. By using RapidEye and aerial digital imagery in a hierarchical supervised classification technique, we were able to produce a high resolution land cover map of Unimak Island. We obtained overall accuracy rates of 71.4 percent which are comparable to other land cover maps using RapidEye imagery. The "red-edge" spectral band included in the RapidEye imagery provides additional spectral information that allows for a more accurate overall classification, raising overall accuracy 5.2 percent.

Analysis of Chi-square Automatic Interaction Detection (CHAID) and Classification and Regression Tree (CRT) for Classification of Corn Production

NASA Astrophysics Data System (ADS)

Susanti, Yuliana; Zukhronah, Etik; Pratiwi, Hasih; Respatiwulan; Sri Sulistijowati, H.

2017-11-01

To achieve food resilience in Indonesia, food diversification by exploring potentials of local food is required. Corn is one of alternating staple food of Javanese society. For that reason, corn production needs to be improved by considering the influencing factors. CHAID and CRT are methods of data mining which can be used to classify the influencing variables. The present study seeks to dig up information on the potentials of local food availability of corn in regencies and cities in Java Island. CHAID analysis yields four classifications with accuracy of 78.8%, while CRT analysis yields seven classifications with accuracy of 79.6%.

Deep multi-scale convolutional neural network for hyperspectral image classification

NASA Astrophysics Data System (ADS)

Zhang, Feng-zhe; Yang, Xia

2018-04-01

In this paper, we proposed a multi-scale convolutional neural network for hyperspectral image classification task. Firstly, compared with conventional convolution, we utilize multi-scale convolutions, which possess larger respective fields, to extract spectral features of hyperspectral image. We design a deep neural network with a multi-scale convolution layer which contains 3 different convolution kernel sizes. Secondly, to avoid overfitting of deep neural network, dropout is utilized, which randomly sleeps neurons, contributing to improve the classification accuracy a bit. In addition, new skills like ReLU in deep learning is utilized in this paper. We conduct experiments on University of Pavia and Salinas datasets, and obtained better classification accuracy compared with other methods.

An evaluation of sampling and full enumeration strategies for Fisher Jenks classification in big data settings

USGS Publications Warehouse

Rey, Sergio J.; Stephens, Philip A.; Laura, Jason R.

2017-01-01

Large data contexts present a number of challenges to optimal choropleth map classifiers. Application of optimal classifiers to a sample of the attribute space is one proposed solution. The properties of alternative sampling-based classification methods are examined through a series of Monte Carlo simulations. The impacts of spatial autocorrelation, number of desired classes, and form of sampling are shown to have significant impacts on the accuracy of map classifications. Tradeoffs between improved speed of the sampling approaches and loss of accuracy are also considered. The results suggest the possibility of guiding the choice of classification scheme as a function of the properties of large data sets.

Proposed hybrid-classifier ensemble algorithm to map snow cover area

NASA Astrophysics Data System (ADS)

Nijhawan, Rahul; Raman, Balasubramanian; Das, Josodhir

2018-01-01

Metaclassification ensemble approach is known to improve the prediction performance of snow-covered area. The methodology adopted in this case is based on neural network along with four state-of-art machine learning algorithms: support vector machine, artificial neural networks, spectral angle mapper, K-mean clustering, and a snow index: normalized difference snow index. An AdaBoost ensemble algorithm related to decision tree for snow-cover mapping is also proposed. According to available literature, these methods have been rarely used for snow-cover mapping. Employing the above techniques, a study was conducted for Raktavarn and Chaturangi Bamak glaciers, Uttarakhand, Himalaya using multispectral Landsat 7 ETM+ (enhanced thematic mapper) image. The study also compares the results with those obtained from statistical combination methods (majority rule and belief functions) and accuracies of individual classifiers. Accuracy assessment is performed by computing the quantity and allocation disagreement, analyzing statistic measures (accuracy, precision, specificity, AUC, and sensitivity) and receiver operating characteristic curves. A total of 225 combinations of parameters for individual classifiers were trained and tested on the dataset and results were compared with the proposed approach. It was observed that the proposed methodology produced the highest classification accuracy (95.21%), close to (94.01%) that was produced by the proposed AdaBoost ensemble algorithm. From the sets of observations, it was concluded that the ensemble of classifiers produced better results compared to individual classifiers.

A new self-report inventory of dyslexia for students: criterion and construct validity.

PubMed

Tamboer, Peter; Vorst, Harrie C M

2015-02-01

The validity of a Dutch self-report inventory of dyslexia was ascertained in two samples of students. Six biographical questions, 20 general language statements and 56 specific language statements were based on dyslexia as a multi-dimensional deficit. Dyslexia and non-dyslexia were assessed with two criteria: identification with test results (Sample 1) and classification using biographical information (both samples). Using discriminant analyses, these criteria were predicted with various groups of statements. All together, 11 discriminant functions were used to estimate classification accuracy of the inventory. In Sample 1, 15 statements predicted the test criterion with classification accuracy of 98%, and 18 statements predicted the biographical criterion with classification accuracy of 97%. In Sample 2, 16 statements predicted the biographical criterion with classification accuracy of 94%. Estimations of positive and negative predictive value were 89% and 99%. Items of various discriminant functions were factor analysed to find characteristic difficulties of students with dyslexia, resulting in a five-factor structure in Sample 1 and a four-factor structure in Sample 2. Answer bias was investigated with measures of internal consistency reliability. Less than 20 self-report items are sufficient to accurately classify students with and without dyslexia. This supports the usefulness of self-assessment of dyslexia as a valid alternative to diagnostic test batteries. Copyright © 2015 John Wiley & Sons, Ltd.

A novel transferable individual tree crown delineation model based on Fishing Net Dragging and boundary classification

NASA Astrophysics Data System (ADS)

Liu, Tao; Im, Jungho; Quackenbush, Lindi J.

2015-12-01

This study provides a novel approach to individual tree crown delineation (ITCD) using airborne Light Detection and Ranging (LiDAR) data in dense natural forests using two main steps: crown boundary refinement based on a proposed Fishing Net Dragging (FiND) method, and segment merging based on boundary classification. FiND starts with approximate tree crown boundaries derived using a traditional watershed method with Gaussian filtering and refines these boundaries using an algorithm that mimics how a fisherman drags a fishing net. Random forest machine learning is then used to classify boundary segments into two classes: boundaries between trees and boundaries between branches that belong to a single tree. Three groups of LiDAR-derived features-two from the pseudo waveform generated along with crown boundaries and one from a canopy height model (CHM)-were used in the classification. The proposed ITCD approach was tested using LiDAR data collected over a mountainous region in the Adirondack Park, NY, USA. Overall accuracy of boundary classification was 82.4%. Features derived from the CHM were generally more important in the classification than the features extracted from the pseudo waveform. A comprehensive accuracy assessment scheme for ITCD was also introduced by considering both area of crown overlap and crown centroids. Accuracy assessment using this new scheme shows the proposed ITCD achieved 74% and 78% as overall accuracy, respectively, for deciduous and mixed forest.

Effects of sample survey design on the accuracy of classification tree models in species distribution models

Treesearch

Thomas C. Edwards; D. Richard Cutler; Niklaus E. Zimmermann; Linda Geiser; Gretchen G. Moisen

2006-01-01

We evaluated the effects of probabilistic (hereafter DESIGN) and non-probabilistic (PURPOSIVE) sample surveys on resultant classification tree models for predicting the presence of four lichen species in the Pacific Northwest, USA. Models derived from both survey forms were assessed using an independent data set (EVALUATION). Measures of accuracy as gauged by...

Using New Models to Analyze Complex Regularities of the World: Commentary on Musso et al. (2013)

ERIC Educational Resources Information Center

Nokelainen, Petri; Silander, Tomi

2014-01-01

This commentary to the recent article by Musso et al. (2013) discusses issues related to model fitting, comparison of classification accuracy of generative and discriminative models, and two (or more) cultures of data modeling. We start by questioning the extremely high classification accuracy with an empirical data from a complex domain. There is…

Predictive Utility and Classification Accuracy of Oral Reading Fluency and the Measures of Academic Progress for the Wisconsin Knowledge and Concepts Exam

ERIC Educational Resources Information Center

Ball, Carrie R.; O'Connor, Edward

2016-01-01

This study examined the predictive validity and classification accuracy of two commonly used universal screening measures relative to a statewide achievement test. Results indicated that second-grade performance on oral reading fluency and the Measures of Academic Progress (MAP), together with special education status, explained 68% of the…

Two Approaches to Estimation of Classification Accuracy Rate under Item Response Theory

ERIC Educational Resources Information Center

Lathrop, Quinn N.; Cheng, Ying

2013-01-01

Within the framework of item response theory (IRT), there are two recent lines of work on the estimation of classification accuracy (CA) rate. One approach estimates CA when decisions are made based on total sum scores, the other based on latent trait estimates. The former is referred to as the Lee approach, and the latter, the Rudner approach,…

Subject-Adaptive Real-Time Sleep Stage Classification Based on Conditional Random Field

PubMed Central

Luo, Gang; Min, Wanli

2007-01-01

Sleep staging is the pattern recognition task of classifying sleep recordings into sleep stages. This task is one of the most important steps in sleep analysis. It is crucial for the diagnosis and treatment of various sleep disorders, and also relates closely to brain-machine interfaces. We report an automatic, online sleep stager using electroencephalogram (EEG) signal based on a recently-developed statistical pattern recognition method, conditional random field, and novel potential functions that have explicit physical meanings. Using sleep recordings from human subjects, we show that the average classification accuracy of our sleep stager almost approaches the theoretical limit and is about 8% higher than that of existing systems. Moreover, for a new subject snew with limited training data Dnew, we perform subject adaptation to improve classification accuracy. Our idea is to use the knowledge learned from old subjects to obtain from Dnew a regulated estimate of CRF’s parameters. Using sleep recordings from human subjects, we show that even without any Dnew, our sleep stager can achieve an average classification accuracy of 70% on snew. This accuracy increases with the size of Dnew and eventually becomes close to the theoretical limit. PMID:18693884

Influence of different topographic correction strategies on mountain vegetation classification accuracy in the Lancang Watershed, China

NASA Astrophysics Data System (ADS)

Zhang, Zhiming; de Wulf, Robert R.; van Coillie, Frieke M. B.; Verbeke, Lieven P. C.; de Clercq, Eva M.; Ou, Xiaokun

2011-01-01

Mapping of vegetation using remote sensing in mountainous areas is considerably hampered by topographic effects on the spectral response pattern. A variety of topographic normalization techniques have been proposed to correct these illumination effects due to topography. The purpose of this study was to compare six different topographic normalization methods (Cosine correction, Minnaert correction, C-correction, Sun-canopy-sensor correction, two-stage topographic normalization, and slope matching technique) for their effectiveness in enhancing vegetation classification in mountainous environments. Since most of the vegetation classes in the rugged terrain of the Lancang Watershed (China) did not feature a normal distribution, artificial neural networks (ANNs) were employed as a classifier. Comparing the ANN classifications, none of the topographic correction methods could significantly improve ETM+ image classification overall accuracy. Nevertheless, at the class level, the accuracy of pine forest could be increased by using topographically corrected images. On the contrary, oak forest and mixed forest accuracies were significantly decreased by using corrected images. The results also showed that none of the topographic normalization strategies was satisfactorily able to correct for the topographic effects in severely shadowed areas.

Support vector machine and principal component analysis for microarray data classification

NASA Astrophysics Data System (ADS)

Astuti, Widi; Adiwijaya

2018-03-01

Cancer is a leading cause of death worldwide although a significant proportion of it can be cured if it is detected early. In recent decades, technology called microarray takes an important role in the diagnosis of cancer. By using data mining technique, microarray data classification can be performed to improve the accuracy of cancer diagnosis compared to traditional techniques. The characteristic of microarray data is small sample but it has huge dimension. Since that, there is a challenge for researcher to provide solutions for microarray data classification with high performance in both accuracy and running time. This research proposed the usage of Principal Component Analysis (PCA) as a dimension reduction method along with Support Vector Method (SVM) optimized by kernel functions as a classifier for microarray data classification. The proposed scheme was applied on seven data sets using 5-fold cross validation and then evaluation and analysis conducted on term of both accuracy and running time. The result showed that the scheme can obtained 100% accuracy for Ovarian and Lung Cancer data when Linear and Cubic kernel functions are used. In term of running time, PCA greatly reduced the running time for every data sets.

Convolutional neural network for high-accuracy functional near-infrared spectroscopy in a brain-computer interface: three-class classification of rest, right-, and left-hand motor execution.

PubMed

Trakoolwilaiwan, Thanawin; Behboodi, Bahareh; Lee, Jaeseok; Kim, Kyungsoo; Choi, Ji-Woong

2018-01-01

The aim of this work is to develop an effective brain-computer interface (BCI) method based on functional near-infrared spectroscopy (fNIRS). In order to improve the performance of the BCI system in terms of accuracy, the ability to discriminate features from input signals and proper classification are desired. Previous studies have mainly extracted features from the signal manually, but proper features need to be selected carefully. To avoid performance degradation caused by manual feature selection, we applied convolutional neural networks (CNNs) as the automatic feature extractor and classifier for fNIRS-based BCI. In this study, the hemodynamic responses evoked by performing rest, right-, and left-hand motor execution tasks were measured on eight healthy subjects to compare performances. Our CNN-based method provided improvements in classification accuracy over conventional methods employing the most commonly used features of mean, peak, slope, variance, kurtosis, and skewness, classified by support vector machine (SVM) and artificial neural network (ANN). Specifically, up to 6.49% and 3.33% improvement in classification accuracy was achieved by CNN compared with SVM and ANN, respectively.

Detecting epileptic seizure with different feature extracting strategies using robust machine learning classification techniques by applying advance parameter optimization approach.

PubMed

Hussain, Lal

2018-06-01

Epilepsy is a neurological disorder produced due to abnormal excitability of neurons in the brain. The research reveals that brain activity is monitored through electroencephalogram (EEG) of patients suffered from seizure to detect the epileptic seizure. The performance of EEG detection based epilepsy require feature extracting strategies. In this research, we have extracted varying features extracting strategies based on time and frequency domain characteristics, nonlinear, wavelet based entropy and few statistical features. A deeper study was undertaken using novel machine learning classifiers by considering multiple factors. The support vector machine kernels are evaluated based on multiclass kernel and box constraint level. Likewise, for K-nearest neighbors (KNN), we computed the different distance metrics, Neighbor weights and Neighbors. Similarly, the decision trees we tuned the paramours based on maximum splits and split criteria and ensemble classifiers are evaluated based on different ensemble methods and learning rate. For training/testing tenfold Cross validation was employed and performance was evaluated in form of TPR, NPR, PPV, accuracy and AUC. In this research, a deeper analysis approach was performed using diverse features extracting strategies using robust machine learning classifiers with more advanced optimal options. Support Vector Machine linear kernel and KNN with City block distance metric give the overall highest accuracy of 99.5% which was higher than using the default parameters for these classifiers. Moreover, highest separation (AUC = 0.9991, 0.9990) were obtained at different kernel scales using SVM. Additionally, the K-nearest neighbors with inverse squared distance weight give higher performance at different Neighbors. Moreover, to distinguish the postictal heart rate oscillations from epileptic ictal subjects, and highest performance of 100% was obtained using different machine learning classifiers.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Sorensen, J; Duran, C; Stingo, F

Purpose: To characterize the effect of virtual monochromatic reconstructions on several commonly used texture analysis features in DECT of the chest. Further, to assess the effect of monochromatic energy levels on the ability of these textural features to identify tissue types. Methods: 20 consecutive patients underwent chest CTs for evaluation of lung nodules using Siemens Somatom Definition Flash DECT. Virtual monochromatic images were constructed at 10keV intervals from 40–190keV. For each patient, an ROI delineated the lesion under investigation, and cylindrical ROI’s were placed within 5 different healthy tissues (blood, fat, muscle, lung, and liver). Several histogram- and Grey Levelmore » Cooccurrence Matrix (GLCM)-based texture features were then evaluated in each ROI at each energy level. As a means of validation, these feature values were then used in a random forest classifier to attempt to identify the tissue types present within each ROI. Their predictive accuracy at each energy level was recorded. Results: All textural features changed considerably with virtual monochromatic energy, particularly below 70keV. Most features exhibited a global minimum or maximum around 80keV, and while feature values changed with energy above this, patient ranking was generally unaffected. As expected, blood demonstrated the lowest inter-patient variability, for all features, while lung lesions (encompassing many different pathologies) exhibited the highest. The accuracy of these features in identifying tissues (76% accuracy) was highest at 80keV, but no clear relationship between energy and classification accuracy was found. Two common misclassifications (blood vs liver and muscle vs fat) accounted for the majority (24 of the 28) errors observed. Conclusion: All textural features were highly dependent on virtual monochromatic energy level, especially below 80keV, and were more stable above this energy. However, in a random forest model, these commonly used features were able to reliably differentiate between most tissues types regardless of energy level. Dr Godoy has received a dual-energy CT research grant from Siemens Healthcare. That grant did not directly fund this research.« less

Regularised extreme learning machine with misclassification cost and rejection cost for gene expression data classification.

PubMed

Lu, Huijuan; Wei, Shasha; Zhou, Zili; Miao, Yanzi; Lu, Yi

2015-01-01

The main purpose of traditional classification algorithms on bioinformatics application is to acquire better classification accuracy. However, these algorithms cannot meet the requirement that minimises the average misclassification cost. In this paper, a new algorithm of cost-sensitive regularised extreme learning machine (CS-RELM) was proposed by using probability estimation and misclassification cost to reconstruct the classification results. By improving the classification accuracy of a group of small sample which higher misclassification cost, the new CS-RELM can minimise the classification cost. The 'rejection cost' was integrated into CS-RELM algorithm to further reduce the average misclassification cost. By using Colon Tumour dataset and SRBCT (Small Round Blue Cells Tumour) dataset, CS-RELM was compared with other cost-sensitive algorithms such as extreme learning machine (ELM), cost-sensitive extreme learning machine, regularised extreme learning machine, cost-sensitive support vector machine (SVM). The results of experiments show that CS-RELM with embedded rejection cost could reduce the average cost of misclassification and made more credible classification decision than others.

Delineation of marsh types of the Texas coast from Corpus Christi Bay to the Sabine River in 2010

USGS Publications Warehouse

Enwright, Nicholas M.; Hartley, Stephen B.; Brasher, Michael G.; Visser, Jenneke M.; Mitchell, Michael K.; Ballard, Bart M.; Parr, Mark W.; Couvillion, Brady R.; Wilson, Barry C.

2014-01-01

Coastal zone managers and researchers often require detailed information regarding emergent marsh vegetation types for modeling habitat capacities and needs of marsh-reliant wildlife (such as waterfowl and alligator). Detailed information on the extent and distribution of marsh vegetation zones throughout the Texas coast has been historically unavailable. In response, the U.S. Geological Survey, in cooperation and collaboration with the U.S. Fish and Wildlife Service via the Gulf Coast Joint Venture, Texas A&M University-Kingsville, the University of Louisiana-Lafayette, and Ducks Unlimited, Inc., has produced a classification of marsh vegetation types along the middle and upper Texas coast from Corpus Christi Bay to the Sabine River. This study incorporates approximately 1,000 ground reference locations collected via helicopter surveys in coastal marsh areas and about 2,000 supplemental locations from fresh marsh, water, and “other” (that is, nonmarsh) areas. About two-thirds of these data were used for training, and about one-third were used for assessing accuracy. Decision-tree analyses using Rulequest See5 were used to classify emergent marsh vegetation types by using these data, multitemporal satellite-based multispectral imagery from 2009 to 2011, a bare-earth digital elevation model (DEM) based on airborne light detection and ranging (lidar), alternative contemporary land cover classifications, and other spatially explicit variables believed to be important for delineating the extent and distribution of marsh vegetation communities. Image objects were generated from segmentation of high-resolution airborne imagery acquired in 2010 and were used to refine the classification. The classification is dated 2010 because the year is both the midpoint of the multitemporal satellite-based imagery (2009–11) classified and the date of the high-resolution airborne imagery that was used to develop image objects. Overall accuracy corrected for bias (accuracy estimate incorporates true marginal proportions) was 91 percent (95 percent confidence interval [CI]: 89.2–92.8), with a kappa statistic of 0.79 (95 percent CI: 0.77–0.81). The classification performed best for saline marsh (user’s accuracy 81.5 percent; producer’s accuracy corrected for bias 62.9 percent) but showed a lesser ability to discriminate intermediate marsh (user’s accuracy 47.7 percent; producer’s accuracy corrected for bias 49.5 percent). Because of confusion in intermediate and brackish marsh classes, an alternative classification containing only three marsh types was created in which intermediate and brackish marshes were combined into a single class. Image objects were reattributed by using this alternative three-marsh-type classification. Overall accuracy, corrected for bias, of this more general classification was 92.4 percent (95 percent CI: 90.7–94.2), and the kappa statistic was 0.83 (95 percent CI: 0.81–0.85). Mean user’s accuracy for marshes within the four-marsh-type and three-marsh-type classifications was 65.4 percent and 75.6 percent, respectively, whereas mean producer’s accuracy was 56.7 percent and 65.1 percent, respectively. This study provides a more objective and repeatable method for classifying marsh types of the middle and upper Texas coast at an extent and greater level of detail than previously available for the study area. The seamless classification produced through this work is now available to help State agencies (such as the Texas Parks and Wildlife Department) and landscape-scale conservation partnerships (such as the Gulf Coast Prairie Landscape Conservation Cooperative and the Gulf Coast Joint Venture) to develop and (or) refine conservation plans targeting priority natural resources. Moreover, these data may improve projections of landscape change and serve as a baseline for monitoring future changes resulting from chronic and episodic stressors.

Increasing accuracy of vehicle detection from conventional vehicle detectors - counts, speeds, classification, and travel time.

DOT National Transportation Integrated Search

2014-09-01

Vehicle classification is an important traffic parameter for transportation planning and infrastructure : management. Length-based vehicle classification from dual loop detectors is among the lowest cost : technologies commonly used for collecting th...

Virtual Sensor of Surface Electromyography in a New Extensive Fault-Tolerant Classification System.

PubMed

de Moura, Karina de O A; Balbinot, Alexandre

2018-05-01

A few prosthetic control systems in the scientific literature obtain pattern recognition algorithms adapted to changes that occur in the myoelectric signal over time and, frequently, such systems are not natural and intuitive. These are some of the several challenges for myoelectric prostheses for everyday use. The concept of the virtual sensor, which has as its fundamental objective to estimate unavailable measures based on other available measures, is being used in other fields of research. The virtual sensor technique applied to surface electromyography can help to minimize these problems, typically related to the degradation of the myoelectric signal that usually leads to a decrease in the classification accuracy of the movements characterized by computational intelligent systems. This paper presents a virtual sensor in a new extensive fault-tolerant classification system to maintain the classification accuracy after the occurrence of the following contaminants: ECG interference, electrode displacement, movement artifacts, power line interference, and saturation. The Time-Varying Autoregressive Moving Average (TVARMA) and Time-Varying Kalman filter (TVK) models are compared to define the most robust model for the virtual sensor. Results of movement classification were presented comparing the usual classification techniques with the method of the degraded signal replacement and classifier retraining. The experimental results were evaluated for these five noise types in 16 surface electromyography (sEMG) channel degradation case studies. The proposed system without using classifier retraining techniques recovered of mean classification accuracy was of 4% to 38% for electrode displacement, movement artifacts, and saturation noise. The best mean classification considering all signal contaminants and channel combinations evaluated was the classification using the retraining method, replacing the degraded channel by the virtual sensor TVARMA model. This method recovered the classification accuracy after the degradations, reaching an average of 5.7% below the classification of the clean signal, that is the signal without the contaminants or the original signal. Moreover, the proposed intelligent technique minimizes the impact of the motion classification caused by signal contamination related to degrading events over time. There are improvements in the virtual sensor model and in the algorithm optimization that need further development to provide an increase the clinical application of myoelectric prostheses but already presents robust results to enable research with virtual sensors on biological signs with stochastic behavior.

Virtual Sensor of Surface Electromyography in a New Extensive Fault-Tolerant Classification System

PubMed Central

Balbinot, Alexandre

2018-01-01

A few prosthetic control systems in the scientific literature obtain pattern recognition algorithms adapted to changes that occur in the myoelectric signal over time and, frequently, such systems are not natural and intuitive. These are some of the several challenges for myoelectric prostheses for everyday use. The concept of the virtual sensor, which has as its fundamental objective to estimate unavailable measures based on other available measures, is being used in other fields of research. The virtual sensor technique applied to surface electromyography can help to minimize these problems, typically related to the degradation of the myoelectric signal that usually leads to a decrease in the classification accuracy of the movements characterized by computational intelligent systems. This paper presents a virtual sensor in a new extensive fault-tolerant classification system to maintain the classification accuracy after the occurrence of the following contaminants: ECG interference, electrode displacement, movement artifacts, power line interference, and saturation. The Time-Varying Autoregressive Moving Average (TVARMA) and Time-Varying Kalman filter (TVK) models are compared to define the most robust model for the virtual sensor. Results of movement classification were presented comparing the usual classification techniques with the method of the degraded signal replacement and classifier retraining. The experimental results were evaluated for these five noise types in 16 surface electromyography (sEMG) channel degradation case studies. The proposed system without using classifier retraining techniques recovered of mean classification accuracy was of 4% to 38% for electrode displacement, movement artifacts, and saturation noise. The best mean classification considering all signal contaminants and channel combinations evaluated was the classification using the retraining method, replacing the degraded channel by the virtual sensor TVARMA model. This method recovered the classification accuracy after the degradations, reaching an average of 5.7% below the classification of the clean signal, that is the signal without the contaminants or the original signal. Moreover, the proposed intelligent technique minimizes the impact of the motion classification caused by signal contamination related to degrading events over time. There are improvements in the virtual sensor model and in the algorithm optimization that need further development to provide an increase the clinical application of myoelectric prostheses but already presents robust results to enable research with virtual sensors on biological signs with stochastic behavior. PMID:29723994

Modeling misregistration and related effects on multispectral classification

NASA Technical Reports Server (NTRS)

Billingsley, F. C.

1981-01-01

The effects of misregistration on the multispectral classification accuracy when the scene registration accuracy is relaxed from 0.3 to 0.5 pixel are investigated. Noise, class separability, spatial transient response, and field size are considered simultaneously with misregistration in their effects on accuracy. Any noise due to the scene, sensor, or to the analog/digital conversion, causes a finite fraction of the measurements to fall outside of the classification limits, even within nominally uniform fields. Misregistration causes field borders in a given band or set of bands to be closer than expected to a given pixel, causing additional pixels to be misclassified due to the mixture of materials in the pixel. Simplified first order models of the various effects are presented, and are used to estimate the performance to be expected.

Automatic classification of hyperactive children: comparing multiple artificial intelligence approaches.

PubMed

Delavarian, Mona; Towhidkhah, Farzad; Gharibzadeh, Shahriar; Dibajnia, Parvin

2011-07-12

Automatic classification of different behavioral disorders with many similarities (e.g. in symptoms) by using an automated approach will help psychiatrists to concentrate on correct disorder and its treatment as soon as possible, to avoid wasting time on diagnosis, and to increase the accuracy of diagnosis. In this study, we tried to differentiate and classify (diagnose) 306 children with many similar symptoms and different behavioral disorders such as ADHD, depression, anxiety, comorbid depression and anxiety and conduct disorder with high accuracy. Classification was based on the symptoms and their severity. With examining 16 different available classifiers, by using "Prtools", we have proposed nearest mean classifier as the most accurate classifier with 96.92% accuracy in this research. Copyright © 2011 Elsevier Ireland Ltd. All rights reserved.

Effective and extensible feature extraction method using genetic algorithm-based frequency-domain feature search for epileptic EEG multiclassification

PubMed Central

Wen, Tingxi; Zhang, Zhongnan

2017-01-01

Abstract In this paper, genetic algorithm-based frequency-domain feature search (GAFDS) method is proposed for the electroencephalogram (EEG) analysis of epilepsy. In this method, frequency-domain features are first searched and then combined with nonlinear features. Subsequently, these features are selected and optimized to classify EEG signals. The extracted features are analyzed experimentally. The features extracted by GAFDS show remarkable independence, and they are superior to the nonlinear features in terms of the ratio of interclass distance and intraclass distance. Moreover, the proposed feature search method can search for features of instantaneous frequency in a signal after Hilbert transformation. The classification results achieved using these features are reasonable; thus, GAFDS exhibits good extensibility. Multiple classical classifiers (i.e., k-nearest neighbor, linear discriminant analysis, decision tree, AdaBoost, multilayer perceptron, and Naïve Bayes) achieve satisfactory classification accuracies by using the features generated by the GAFDS method and the optimized feature selection. The accuracies for 2-classification and 3-classification problems may reach up to 99% and 97%, respectively. Results of several cross-validation experiments illustrate that GAFDS is effective in the extraction of effective features for EEG classification. Therefore, the proposed feature selection and optimization model can improve classification accuracy. PMID:28489789

Effective and extensible feature extraction method using genetic algorithm-based frequency-domain feature search for epileptic EEG multiclassification.

PubMed

Wen, Tingxi; Zhang, Zhongnan

2017-05-01

In this paper, genetic algorithm-based frequency-domain feature search (GAFDS) method is proposed for the electroencephalogram (EEG) analysis of epilepsy. In this method, frequency-domain features are first searched and then combined with nonlinear features. Subsequently, these features are selected and optimized to classify EEG signals. The extracted features are analyzed experimentally. The features extracted by GAFDS show remarkable independence, and they are superior to the nonlinear features in terms of the ratio of interclass distance and intraclass distance. Moreover, the proposed feature search method can search for features of instantaneous frequency in a signal after Hilbert transformation. The classification results achieved using these features are reasonable; thus, GAFDS exhibits good extensibility. Multiple classical classifiers (i.e., k-nearest neighbor, linear discriminant analysis, decision tree, AdaBoost, multilayer perceptron, and Naïve Bayes) achieve satisfactory classification accuracies by using the features generated by the GAFDS method and the optimized feature selection. The accuracies for 2-classification and 3-classification problems may reach up to 99% and 97%, respectively. Results of several cross-validation experiments illustrate that GAFDS is effective in the extraction of effective features for EEG classification. Therefore, the proposed feature selection and optimization model can improve classification accuracy.

Land cover classification accuracy from electro-optical, X, C, and L-band Synthetic Aperture Radar data fusion

NASA Astrophysics Data System (ADS)

Hammann, Mark Gregory

The fusion of electro-optical (EO) multi-spectral satellite imagery with Synthetic Aperture Radar (SAR) data was explored with the working hypothesis that the addition of multi-band SAR will increase the land-cover (LC) classification accuracy compared to EO alone. Three satellite sources for SAR imagery were used: X-band from TerraSAR-X, C-band from RADARSAT-2, and L-band from PALSAR. Images from the RapidEye satellites were the source of the EO imagery. Imagery from the GeoEye-1 and WorldView-2 satellites aided the selection of ground truth. Three study areas were chosen: Wad Medani, Sudan; Campinas, Brazil; and Fresno- Kings Counties, USA. EO imagery were radiometrically calibrated, atmospherically compensated, orthorectifed, co-registered, and clipped to a common area of interest (AOI). SAR imagery were radiometrically calibrated, and geometrically corrected for terrain and incidence angle by converting to ground range and Sigma Naught (?0). The original SAR HH data were included in the fused image stack after despeckling with a 3x3 Enhanced Lee filter. The variance and Gray-Level-Co-occurrence Matrix (GLCM) texture measures of contrast, entropy, and correlation were derived from the non-despeckled SAR HH bands. Data fusion was done with layer stacking and all data were resampled to a common spatial resolution. The Support Vector Machine (SVM) decision rule was used for the supervised classifications. Similar LC classes were identified and tested for each study area. For Wad Medani, nine classes were tested: low and medium intensity urban, sparse forest, water, barren ground, and four agriculture classes (fallow, bare agricultural ground, green crops, and orchards). For Campinas, Brazil, five generic classes were tested: urban, agriculture, forest, water, and barren ground. For the Fresno-Kings Counties location 11 classes were studied: three generic classes (urban, water, barren land), and eight specific crops. In all cases the addition of SAR to EO resulted in higher overall classification accuracies. In many cases using more than a single SAR band also improved the classification accuracy. There was no single best SAR band for all cases; for specific study areas or LC classes, different SAR bands were better. For Wad Medani, the overall accuracy increased nearly 25% over EO by using all three SAR bands and GLCM texture. For Campinas, the improvement over EO was 4.3%; the large areas of vegetation were classified by EO with good accuracy. At Fresno-Kings Counties, EO+SAR fusion improved the overall classification accuracy by 7%. For times or regions where EO is not available due to extended cloud cover, classification with SAR is often the only option; note that SAR alone typically results in lower classification accuracies than when using EO or EO-SAR fusion. Fusion of EO and SAR was especially important to improve the separability of orchards from other crops, and separating urban areas with buildings from bare soil; those classes are difficult to accurately separate with EO. The outcome of this dissertation contributes to the understanding of the benefits of combining data from EO imagery with different SAR bands and SAR derived texture data to identify different LC classes. In times of increased public and private budget constraints and industry consolidation, this dissertation provides insight as to which band packages could be most useful for increased accuracy in LC classification.

Fuzzy membership functions for analysis of high-resolution CT images of diffuse pulmonary diseases.

PubMed

Almeida, Eliana; Rangayyan, Rangaraj M; Azevedo-Marques, Paulo M

2015-08-01

We propose the use of fuzzy membership functions to analyze images of diffuse pulmonary diseases (DPDs) based on fractal and texture features. The features were extracted from preprocessed regions of interest (ROIs) selected from high-resolution computed tomography images. The ROIs represent five different patterns of DPDs and normal lung tissue. A Gaussian mixture model (GMM) was constructed for each feature, with six Gaussians modeling the six patterns. Feature selection was performed and the GMMs of the five significant features were used. From the GMMs, fuzzy membership functions were obtained by a probability-possibility transformation and further statistical analysis was performed. An average classification accuracy of 63.5% was obtained for the six classes. For four of the six classes, the classification accuracy was superior to 65%, and the best classification accuracy was 75.5% for one class. The use of fuzzy membership functions to assist in pattern classification is an alternative to deterministic approaches to explore strategies for medical diagnosis.

Pathological brain detection based on wavelet entropy and Hu moment invariants.

PubMed

Zhang, Yudong; Wang, Shuihua; Sun, Ping; Phillips, Preetha

2015-01-01

With the aim of developing an accurate pathological brain detection system, we proposed a novel automatic computer-aided diagnosis (CAD) to detect pathological brains from normal brains obtained by magnetic resonance imaging (MRI) scanning. The problem still remained a challenge for technicians and clinicians, since MR imaging generated an exceptionally large information dataset. A new two-step approach was proposed in this study. We used wavelet entropy (WE) and Hu moment invariants (HMI) for feature extraction, and the generalized eigenvalue proximal support vector machine (GEPSVM) for classification. To further enhance classification accuracy, the popular radial basis function (RBF) kernel was employed. The 10 runs of k-fold stratified cross validation result showed that the proposed "WE + HMI + GEPSVM + RBF" method was superior to existing methods w.r.t. classification accuracy. It obtained the average classification accuracies of 100%, 100%, and 99.45% over Dataset-66, Dataset-160, and Dataset-255, respectively. The proposed method is effective and can be applied to realistic use.

MULTIMODAL CLASSIFICATION OF DEMENTIA USING FUNCTIONAL DATA, ANATOMICAL FEATURES AND 3D INVARIANT SHAPE DESCRIPTORS

PubMed Central

Mikhno, Arthur; Nuevo, Pablo Martinez; Devanand, Davangere P.; Parsey, Ramin V.; Laine, Andrew F.

2013-01-01

Multimodality classification of Alzheimer’s disease (AD) and its prodromal stage, Mild Cognitive Impairment (MCI), is of interest to the medical community. We improve on prior classification frameworks by incorporating multiple features from MRI and PET data obtained with multiple radioligands, fluorodeoxyglucose (FDG) and Pittsburg compound B (PIB). We also introduce a new MRI feature, invariant shape descriptors based on 3D Zernike moments applied to the hippocampus region. Classification performance is evaluated on data from 17 healthy controls (CTR), 22 MCI, and 17 AD subjects. Zernike significantly outperforms volume, accuracy (Zernike to volume): CTR/AD (90.7% to 71.6%), CTR/MCI (76.2% to 60.0%), MCI/AD (84.3% to 65.5%). Zernike also provides comparable and complementary performance to PET. Optimal accuracy is achieved when Zernike and PET features are combined (accuracy, specificity, sensitivity), CTR/AD (98.8%, 99.5%, 98.1%), CTR/MCI (84.3%, 82.9%, 85.9%) and MCI/AD (93.3%, 93.6%, 93.3%). PMID:24576927

MULTIMODAL CLASSIFICATION OF DEMENTIA USING FUNCTIONAL DATA, ANATOMICAL FEATURES AND 3D INVARIANT SHAPE DESCRIPTORS.

PubMed

Mikhno, Arthur; Nuevo, Pablo Martinez; Devanand, Davangere P; Parsey, Ramin V; Laine, Andrew F

2012-01-01

Multimodality classification of Alzheimer's disease (AD) and its prodromal stage, Mild Cognitive Impairment (MCI), is of interest to the medical community. We improve on prior classification frameworks by incorporating multiple features from MRI and PET data obtained with multiple radioligands, fluorodeoxyglucose (FDG) and Pittsburg compound B (PIB). We also introduce a new MRI feature, invariant shape descriptors based on 3D Zernike moments applied to the hippocampus region. Classification performance is evaluated on data from 17 healthy controls (CTR), 22 MCI, and 17 AD subjects. Zernike significantly outperforms volume, accuracy (Zernike to volume): CTR/AD (90.7% to 71.6%), CTR/MCI (76.2% to 60.0%), MCI/AD (84.3% to 65.5%). Zernike also provides comparable and complementary performance to PET. Optimal accuracy is achieved when Zernike and PET features are combined (accuracy, specificity, sensitivity), CTR/AD (98.8%, 99.5%, 98.1%), CTR/MCI (84.3%, 82.9%, 85.9%) and MCI/AD (93.3%, 93.6%, 93.3%).

A neural network approach for enhancing information extraction from multispectral image data

USGS Publications Warehouse

Liu, J.; Shao, G.; Zhu, H.; Liu, S.

2005-01-01

A back-propagation artificial neural network (ANN) was applied to classify multispectral remote sensing imagery data. The classification procedure included four steps: (i) noisy training that adds minor random variations to the sampling data to make the data more representative and to reduce the training sample size; (ii) iterative or multi-tier classification that reclassifies the unclassified pixels by making a subset of training samples from the original training set, which means the neural model can focus on fewer classes; (iii) spectral channel selection based on neural network weights that can distinguish the relative importance of each channel in the classification process to simplify the ANN model; and (iv) voting rules that adjust the accuracy of classification and produce outputs of different confidence levels. The Purdue Forest, located west of Purdue University, West Lafayette, Indiana, was chosen as the test site. The 1992 Landsat thematic mapper imagery was used as the input data. High-quality airborne photographs of the same Lime period were used for the ground truth. A total of 11 land use and land cover classes were defined, including water, broadleaved forest, coniferous forest, young forest, urban and road, and six types of cropland-grassland. The experiment, indicated that the back-propagation neural network application was satisfactory in distinguishing different land cover types at US Geological Survey levels II-III. The single-tier classification reached an overall accuracy of 85%. and the multi-tier classification an overall accuracy of 95%. For the whole test, region, the final output of this study reached an overall accuracy of 87%. ?? 2005 CASI.

Multi-class computational evolution: development, benchmark evaluation and application to RNA-Seq biomarker discovery.

PubMed

Crabtree, Nathaniel M; Moore, Jason H; Bowyer, John F; George, Nysia I

2017-01-01

A computational evolution system (CES) is a knowledge discovery engine that can identify subtle, synergistic relationships in large datasets. Pareto optimization allows CESs to balance accuracy with model complexity when evolving classifiers. Using Pareto optimization, a CES is able to identify a very small number of features while maintaining high classification accuracy. A CES can be designed for various types of data, and the user can exploit expert knowledge about the classification problem in order to improve discrimination between classes. These characteristics give CES an advantage over other classification and feature selection algorithms, particularly when the goal is to identify a small number of highly relevant, non-redundant biomarkers. Previously, CESs have been developed only for binary class datasets. In this study, we developed a multi-class CES. The multi-class CES was compared to three common feature selection and classification algorithms: support vector machine (SVM), random k-nearest neighbor (RKNN), and random forest (RF). The algorithms were evaluated on three distinct multi-class RNA sequencing datasets. The comparison criteria were run-time, classification accuracy, number of selected features, and stability of selected feature set (as measured by the Tanimoto distance). The performance of each algorithm was data-dependent. CES performed best on the dataset with the smallest sample size, indicating that CES has a unique advantage since the accuracy of most classification methods suffer when sample size is small. The multi-class extension of CES increases the appeal of its application to complex, multi-class datasets in order to identify important biomarkers and features.

A novel artificial immune clonal selection classification and rule mining with swarm learning model

NASA Astrophysics Data System (ADS)

Al-Sheshtawi, Khaled A.; Abdul-Kader, Hatem M.; Elsisi, Ashraf B.

2013-06-01

Metaheuristic optimisation algorithms have become popular choice for solving complex problems. By integrating Artificial Immune clonal selection algorithm (CSA) and particle swarm optimisation (PSO) algorithm, a novel hybrid Clonal Selection Classification and Rule Mining with Swarm Learning Algorithm (CS2) is proposed. The main goal of the approach is to exploit and explore the parallel computation merit of Clonal Selection and the speed and self-organisation merits of Particle Swarm by sharing information between clonal selection population and particle swarm. Hence, we employed the advantages of PSO to improve the mutation mechanism of the artificial immune CSA and to mine classification rules within datasets. Consequently, our proposed algorithm required less training time and memory cells in comparison to other AIS algorithms. In this paper, classification rule mining has been modelled as a miltiobjective optimisation problem with predictive accuracy. The multiobjective approach is intended to allow the PSO algorithm to return an approximation to the accuracy and comprehensibility border, containing solutions that are spread across the border. We compared our proposed algorithm classification accuracy CS2 with five commonly used CSAs, namely: AIRS1, AIRS2, AIRS-Parallel, CLONALG, and CSCA using eight benchmark datasets. We also compared our proposed algorithm classification accuracy CS2 with other five methods, namely: Naïve Bayes, SVM, MLP, CART, and RFB. The results show that the proposed algorithm is comparable to the 10 studied algorithms. As a result, the hybridisation, built of CSA and PSO, can develop respective merit, compensate opponent defect, and make search-optimal effect and speed better.

Classification of drug molecules considering their IC50 values using mixed-integer linear programming based hyper-boxes method.

PubMed

Armutlu, Pelin; Ozdemir, Muhittin E; Uney-Yuksektepe, Fadime; Kavakli, I Halil; Turkay, Metin

2008-10-03

A priori analysis of the activity of drugs on the target protein by computational approaches can be useful in narrowing down drug candidates for further experimental tests. Currently, there are a large number of computational methods that predict the activity of drugs on proteins. In this study, we approach the activity prediction problem as a classification problem and, we aim to improve the classification accuracy by introducing an algorithm that combines partial least squares regression with mixed-integer programming based hyper-boxes classification method, where drug molecules are classified as low active or high active regarding their binding activity (IC50 values) on target proteins. We also aim to determine the most significant molecular descriptors for the drug molecules. We first apply our approach by analyzing the activities of widely known inhibitor datasets including Acetylcholinesterase (ACHE), Benzodiazepine Receptor (BZR), Dihydrofolate Reductase (DHFR), Cyclooxygenase-2 (COX-2) with known IC50 values. The results at this stage proved that our approach consistently gives better classification accuracies compared to 63 other reported classification methods such as SVM, Naïve Bayes, where we were able to predict the experimentally determined IC50 values with a worst case accuracy of 96%. To further test applicability of this approach we first created dataset for Cytochrome P450 C17 inhibitors and then predicted their activities with 100% accuracy. Our results indicate that this approach can be utilized to predict the inhibitory effects of inhibitors based on their molecular descriptors. This approach will not only enhance drug discovery process, but also save time and resources committed.