A comparison of rule-based and machine learning approaches for classifying patient portal messages.

PubMed

Cronin, Robert M; Fabbri, Daniel; Denny, Joshua C; Rosenbloom, S Trent; Jackson, Gretchen Purcell

2017-09-01

Secure messaging through patient portals is an increasingly popular way that consumers interact with healthcare providers. The increasing burden of secure messaging can affect clinic staffing and workflows. Manual management of portal messages is costly and time consuming. Automated classification of portal messages could potentially expedite message triage and delivery of care. We developed automated patient portal message classifiers with rule-based and machine learning techniques using bag of words and natural language processing (NLP) approaches. To evaluate classifier performance, we used a gold standard of 3253 portal messages manually categorized using a taxonomy of communication types (i.e., main categories of informational, medical, logistical, social, and other communications, and subcategories including prescriptions, appointments, problems, tests, follow-up, contact information, and acknowledgement). We evaluated our classifiers' accuracies in identifying individual communication types within portal messages with area under the receiver-operator curve (AUC). Portal messages often contain more than one type of communication. To predict all communication types within single messages, we used the Jaccard Index. We extracted the variables of importance for the random forest classifiers. The best performing approaches to classification for the major communication types were: logistic regression for medical communications (AUC: 0.899); basic (rule-based) for informational communications (AUC: 0.842); and random forests for social communications and logistical communications (AUCs: 0.875 and 0.925, respectively). The best performing classification approach of classifiers for individual communication subtypes was random forests for Logistical-Contact Information (AUC: 0.963). The Jaccard Indices by approach were: basic classifier, Jaccard Index: 0.674; Naïve Bayes, Jaccard Index: 0.799; random forests, Jaccard Index: 0.859; and logistic regression, Jaccard Index: 0.861. For medical communications, the most predictive variables were NLP concepts (e.g., Temporal_Concept, which maps to 'morning', 'evening' and Idea_or_Concept which maps to 'appointment' and 'refill'). For logistical communications, the most predictive variables contained similar numbers of NLP variables and words (e.g., Telephone mapping to 'phone', 'insurance'). For social and informational communications, the most predictive variables were words (e.g., social: 'thanks', 'much', informational: 'question', 'mean'). This study applies automated classification methods to the content of patient portal messages and evaluates the application of NLP techniques on consumer communications in patient portal messages. We demonstrated that random forest and logistic regression approaches accurately classified the content of portal messages, although the best approach to classification varied by communication type. Words were the most predictive variables for classification of most communication types, although NLP variables were most predictive for medical communication types. As adoption of patient portals increases, automated techniques could assist in understanding and managing growing volumes of messages. Further work is needed to improve classification performance to potentially support message triage and answering. Copyright © 2017 Elsevier B.V. All rights reserved.

Advanced Subspace Techniques for Modeling Channel and Session Variability in a Speaker Recognition System

DTIC Science & Technology

2012-03-01

with each SVM discriminating between a pair of the N total speakers in the data set. The (( + 1))/2 classifiers then vote on the final...classification of a test sample. The Random Forest classifier is an ensemble classifier that votes amongst decision trees generated with each node using...Forest vote , and the effects of overtraining will be mitigated by the fact that each decision tree is overtrained differently (due to the random

Ensemble of random forests One vs. Rest classifiers for MCI and AD prediction using ANOVA cortical and subcortical feature selection and partial least squares.

PubMed

Ramírez, J; Górriz, J M; Ortiz, A; Martínez-Murcia, F J; Segovia, F; Salas-Gonzalez, D; Castillo-Barnes, D; Illán, I A; Puntonet, C G

2018-05-15

Alzheimer's disease (AD) is the most common cause of dementia in the elderly and affects approximately 30 million individuals worldwide. Mild cognitive impairment (MCI) is very frequently a prodromal phase of AD, and existing studies have suggested that people with MCI tend to progress to AD at a rate of about 10-15% per year. However, the ability of clinicians and machine learning systems to predict AD based on MRI biomarkers at an early stage is still a challenging problem that can have a great impact in improving treatments. The proposed system, developed by the SiPBA-UGR team for this challenge, is based on feature standardization, ANOVA feature selection, partial least squares feature dimension reduction and an ensemble of One vs. Rest random forest classifiers. With the aim of improving its performance when discriminating healthy controls (HC) from MCI, a second binary classification level was introduced that reconsiders the HC and MCI predictions of the first level. The system was trained and evaluated on an ADNI datasets that consist of T1-weighted MRI morphological measurements from HC, stable MCI, converter MCI and AD subjects. The proposed system yields a 56.25% classification score on the test subset which consists of 160 real subjects. The classifier yielded the best performance when compared to: (i) One vs. One (OvO), One vs. Rest (OvR) and error correcting output codes (ECOC) as strategies for reducing the multiclass classification task to multiple binary classification problems, (ii) support vector machines, gradient boosting classifier and random forest as base binary classifiers, and (iii) bagging ensemble learning. A robust method has been proposed for the international challenge on MCI prediction based on MRI data. The system yielded the second best performance during the competition with an accuracy rate of 56.25% when evaluated on the real subjects of the test set. Copyright © 2017 Elsevier B.V. All rights reserved.

Object-based forest classification to facilitate landscape-scale conservation in the Mississippi Alluvial Valley

USGS Publications Warehouse

Mitchell, Michael; Wilson, R. Randy; Twedt, Daniel J.; Mini, Anne E.; James, J. Dale

2016-01-01

The Mississippi Alluvial Valley is a floodplain along the southern extent of the Mississippi River extending from southern Missouri to the Gulf of Mexico. This area once encompassed nearly 10 million ha of floodplain forests, most of which has been converted to agriculture over the past two centuries. Conservation programs in this region revolve around protection of existing forest and reforestation of converted lands. Therefore, an accurate and up to date classification of forest cover is essential for conservation planning, including efforts that prioritize areas for conservation activities. We used object-based image analysis with Random Forest classification to quickly and accurately classify forest cover. We used Landsat band, band ratio, and band index statistics to identify and define similar objects as our training sets instead of selecting individual training points. This provided a single rule-set that was used to classify each of the 11 Landsat 5 Thematic Mapper scenes that encompassed the Mississippi Alluvial Valley. We classified 3,307,910±85,344 ha (32% of this region) as forest. Our overall classification accuracy was 96.9% with Kappa statistic of 0.96. Because this method of forest classification is rapid and accurate, assessment of forest cover can be regularly updated and progress toward forest habitat goals identified in conservation plans can be periodically evaluated.

A Machine Learning and Cross-Validation Approach for the Discrimination of Vegetation Physiognomic Types Using Satellite Based Multispectral and Multitemporal Data.

PubMed

Sharma, Ram C; Hara, Keitarou; Hirayama, Hidetake

2017-01-01

This paper presents the performance and evaluation of a number of machine learning classifiers for the discrimination between the vegetation physiognomic classes using the satellite based time-series of the surface reflectance data. Discrimination of six vegetation physiognomic classes, Evergreen Coniferous Forest, Evergreen Broadleaf Forest, Deciduous Coniferous Forest, Deciduous Broadleaf Forest, Shrubs, and Herbs, was dealt with in the research. Rich-feature data were prepared from time-series of the satellite data for the discrimination and cross-validation of the vegetation physiognomic types using machine learning approach. A set of machine learning experiments comprised of a number of supervised classifiers with different model parameters was conducted to assess how the discrimination of vegetation physiognomic classes varies with classifiers, input features, and ground truth data size. The performance of each experiment was evaluated by using the 10-fold cross-validation method. Experiment using the Random Forests classifier provided highest overall accuracy (0.81) and kappa coefficient (0.78). However, accuracy metrics did not vary much with experiments. Accuracy metrics were found to be very sensitive to input features and size of ground truth data. The results obtained in the research are expected to be useful for improving the vegetation physiognomic mapping in Japan.

Neonatal Seizure Detection Using Deep Convolutional Neural Networks.

PubMed

Ansari, Amir H; Cherian, Perumpillichira J; Caicedo, Alexander; Naulaers, Gunnar; De Vos, Maarten; Van Huffel, Sabine

2018-04-02

Identifying a core set of features is one of the most important steps in the development of an automated seizure detector. In most of the published studies describing features and seizure classifiers, the features were hand-engineered, which may not be optimal. The main goal of the present paper is using deep convolutional neural networks (CNNs) and random forest to automatically optimize feature selection and classification. The input of the proposed classifier is raw multi-channel EEG and the output is the class label: seizure/nonseizure. By training this network, the required features are optimized, while fitting a nonlinear classifier on the features. After training the network with EEG recordings of 26 neonates, five end layers performing the classification were replaced with a random forest classifier in order to improve the performance. This resulted in a false alarm rate of 0.9 per hour and seizure detection rate of 77% using a test set of EEG recordings of 22 neonates that also included dubious seizures. The newly proposed CNN classifier outperformed three data-driven feature-based approaches and performed similar to a previously developed heuristic method.

AUTOCLASSIFICATION OF THE VARIABLE 3XMM SOURCES USING THE RANDOM FOREST MACHINE LEARNING ALGORITHM

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

Farrell, Sean A.; Murphy, Tara; Lo, Kitty K., E-mail: s.farrell@physics.usyd.edu.au

In the current era of large surveys and massive data sets, autoclassification of astrophysical sources using intelligent algorithms is becoming increasingly important. In this paper we present the catalog of variable sources in the Third XMM-Newton Serendipitous Source catalog (3XMM) autoclassified using the Random Forest machine learning algorithm. We used a sample of manually classified variable sources from the second data release of the XMM-Newton catalogs (2XMMi-DR2) to train the classifier, obtaining an accuracy of ∼92%. We also evaluated the effectiveness of identifying spurious detections using a sample of spurious sources, achieving an accuracy of ∼95%. Manual investigation of amore » random sample of classified sources confirmed these accuracy levels and showed that the Random Forest machine learning algorithm is highly effective at automatically classifying 3XMM sources. Here we present the catalog of classified 3XMM variable sources. We also present three previously unidentified unusual sources that were flagged as outlier sources by the algorithm: a new candidate supergiant fast X-ray transient, a 400 s X-ray pulsar, and an eclipsing 5 hr binary system coincident with a known Cepheid.« less

Application of lifting wavelet and random forest in compound fault diagnosis of gearbox

NASA Astrophysics Data System (ADS)

Chen, Tang; Cui, Yulian; Feng, Fuzhou; Wu, Chunzhi

2018-03-01

Aiming at the weakness of compound fault characteristic signals of a gearbox of an armored vehicle and difficult to identify fault types, a fault diagnosis method based on lifting wavelet and random forest is proposed. First of all, this method uses the lifting wavelet transform to decompose the original vibration signal in multi-layers, reconstructs the multi-layer low-frequency and high-frequency components obtained by the decomposition to get multiple component signals. Then the time-domain feature parameters are obtained for each component signal to form multiple feature vectors, which is input into the random forest pattern recognition classifier to determine the compound fault type. Finally, a variety of compound fault data of the gearbox fault analog test platform are verified, the results show that the recognition accuracy of the fault diagnosis method combined with the lifting wavelet and the random forest is up to 99.99%.

An assessment of the effectiveness of a random forest classifier for land-cover classification

NASA Astrophysics Data System (ADS)

Rodriguez-Galiano, V. F.; Ghimire, B.; Rogan, J.; Chica-Olmo, M.; Rigol-Sanchez, J. P.

2012-01-01

Land cover monitoring using remotely sensed data requires robust classification methods which allow for the accurate mapping of complex land cover and land use categories. Random forest (RF) is a powerful machine learning classifier that is relatively unknown in land remote sensing and has not been evaluated thoroughly by the remote sensing community compared to more conventional pattern recognition techniques. Key advantages of RF include: their non-parametric nature; high classification accuracy; and capability to determine variable importance. However, the split rules for classification are unknown, therefore RF can be considered to be black box type classifier. RF provides an algorithm for estimating missing values; and flexibility to perform several types of data analysis, including regression, classification, survival analysis, and unsupervised learning. In this paper, the performance of the RF classifier for land cover classification of a complex area is explored. Evaluation was based on several criteria: mapping accuracy, sensitivity to data set size and noise. Landsat-5 Thematic Mapper data captured in European spring and summer were used with auxiliary variables derived from a digital terrain model to classify 14 different land categories in the south of Spain. Results show that the RF algorithm yields accurate land cover classifications, with 92% overall accuracy and a Kappa index of 0.92. RF is robust to training data reduction and noise because significant differences in kappa values were only observed for data reduction and noise addition values greater than 50 and 20%, respectively. Additionally, variables that RF identified as most important for classifying land cover coincided with expectations. A McNemar test indicates an overall better performance of the random forest model over a single decision tree at the 0.00001 significance level.

Tissue segmentation of computed tomography images using a Random Forest algorithm: a feasibility study

NASA Astrophysics Data System (ADS)

Polan, Daniel F.; Brady, Samuel L.; Kaufman, Robert A.

2016-09-01

There is a need for robust, fully automated whole body organ segmentation for diagnostic CT. This study investigates and optimizes a Random Forest algorithm for automated organ segmentation; explores the limitations of a Random Forest algorithm applied to the CT environment; and demonstrates segmentation accuracy in a feasibility study of pediatric and adult patients. To the best of our knowledge, this is the first study to investigate a trainable Weka segmentation (TWS) implementation using Random Forest machine-learning as a means to develop a fully automated tissue segmentation tool developed specifically for pediatric and adult examinations in a diagnostic CT environment. Current innovation in computed tomography (CT) is focused on radiomics, patient-specific radiation dose calculation, and image quality improvement using iterative reconstruction, all of which require specific knowledge of tissue and organ systems within a CT image. The purpose of this study was to develop a fully automated Random Forest classifier algorithm for segmentation of neck-chest-abdomen-pelvis CT examinations based on pediatric and adult CT protocols. Seven materials were classified: background, lung/internal air or gas, fat, muscle, solid organ parenchyma, blood/contrast enhanced fluid, and bone tissue using Matlab and the TWS plugin of FIJI. The following classifier feature filters of TWS were investigated: minimum, maximum, mean, and variance evaluated over a voxel radius of 2 n , (n from 0 to 4), along with noise reduction and edge preserving filters: Gaussian, bilateral, Kuwahara, and anisotropic diffusion. The Random Forest algorithm used 200 trees with 2 features randomly selected per node. The optimized auto-segmentation algorithm resulted in 16 image features including features derived from maximum, mean, variance Gaussian and Kuwahara filters. Dice similarity coefficient (DSC) calculations between manually segmented and Random Forest algorithm segmented images from 21 patient image sections, were analyzed. The automated algorithm produced segmentation of seven material classes with a median DSC of 0.86  ±  0.03 for pediatric patient protocols, and 0.85  ±  0.04 for adult patient protocols. Additionally, 100 randomly selected patient examinations were segmented and analyzed, and a mean sensitivity of 0.91 (range: 0.82-0.98), specificity of 0.89 (range: 0.70-0.98), and accuracy of 0.90 (range: 0.76-0.98) were demonstrated. In this study, we demonstrate that this fully automated segmentation tool was able to produce fast and accurate segmentation of the neck and trunk of the body over a wide range of patient habitus and scan parameters.

Evaluation of Semi-supervised Learning for Classification of Protein Crystallization Imagery.

PubMed

Sigdel, Madhav; Dinç, İmren; Dinç, Semih; Sigdel, Madhu S; Pusey, Marc L; Aygün, Ramazan S

2014-03-01

In this paper, we investigate the performance of two wrapper methods for semi-supervised learning algorithms for classification of protein crystallization images with limited labeled images. Firstly, we evaluate the performance of semi-supervised approach using self-training with naïve Bayesian (NB) and sequential minimum optimization (SMO) as the base classifiers. The confidence values returned by these classifiers are used to select high confident predictions to be used for self-training. Secondly, we analyze the performance of Yet Another Two Stage Idea (YATSI) semi-supervised learning using NB, SMO, multilayer perceptron (MLP), J48 and random forest (RF) classifiers. These results are compared with the basic supervised learning using the same training sets. We perform our experiments on a dataset consisting of 2250 protein crystallization images for different proportions of training and test data. Our results indicate that NB and SMO using both self-training and YATSI semi-supervised approaches improve accuracies with respect to supervised learning. On the other hand, MLP, J48 and RF perform better using basic supervised learning. Overall, random forest classifier yields the best accuracy with supervised learning for our dataset.

Introducing two Random Forest based methods for cloud detection in remote sensing images

NASA Astrophysics Data System (ADS)

Ghasemian, Nafiseh; Akhoondzadeh, Mehdi

2018-07-01

Cloud detection is a necessary phase in satellite images processing to retrieve the atmospheric and lithospheric parameters. Currently, some cloud detection methods based on Random Forest (RF) model have been proposed but they do not consider both spectral and textural characteristics of the image. Furthermore, they have not been tested in the presence of snow/ice. In this paper, we introduce two RF based algorithms, Feature Level Fusion Random Forest (FLFRF) and Decision Level Fusion Random Forest (DLFRF) to incorporate visible, infrared (IR) and thermal spectral and textural features (FLFRF) including Gray Level Co-occurrence Matrix (GLCM) and Robust Extended Local Binary Pattern (RELBP_CI) or visible, IR and thermal classifiers (DLFRF) for highly accurate cloud detection on remote sensing images. FLFRF first fuses visible, IR and thermal features. Thereafter, it uses the RF model to classify pixels to cloud, snow/ice and background or thick cloud, thin cloud and background. DLFRF considers visible, IR and thermal features (both spectral and textural) separately and inserts each set of features to RF model. Then, it holds vote matrix of each run of the model. Finally, it fuses the classifiers using the majority vote method. To demonstrate the effectiveness of the proposed algorithms, 10 Terra MODIS and 15 Landsat 8 OLI/TIRS images with different spatial resolutions are used in this paper. Quantitative analyses are based on manually selected ground truth data. Results show that after adding RELBP_CI to input feature set cloud detection accuracy improves. Also, the average cloud kappa values of FLFRF and DLFRF on MODIS images (1 and 0.99) are higher than other machine learning methods, Linear Discriminate Analysis (LDA), Classification And Regression Tree (CART), K Nearest Neighbor (KNN) and Support Vector Machine (SVM) (0.96). The average snow/ice kappa values of FLFRF and DLFRF on MODIS images (1 and 0.85) are higher than other traditional methods. The quantitative values on Landsat 8 images show similar trend. Consequently, while SVM and K-nearest neighbor show overestimation in predicting cloud and snow/ice pixels, our Random Forest (RF) based models can achieve higher cloud, snow/ice kappa values on MODIS and thin cloud, thick cloud and snow/ice kappa values on Landsat 8 images. Our algorithms predict both thin and thick cloud on Landsat 8 images while the existing cloud detection algorithm, Fmask cannot discriminate them. Compared to the state-of-the-art methods, our algorithms have acquired higher average cloud and snow/ice kappa values for different spatial resolutions.

A Dirichlet-Multinomial Bayes Classifier for Disease Diagnosis with Microbial Compositions.

PubMed

Gao, Xiang; Lin, Huaiying; Dong, Qunfeng

2017-01-01

Dysbiosis of microbial communities is associated with various human diseases, raising the possibility of using microbial compositions as biomarkers for disease diagnosis. We have developed a Bayes classifier by modeling microbial compositions with Dirichlet-multinomial distributions, which are widely used to model multicategorical count data with extra variation. The parameters of the Dirichlet-multinomial distributions are estimated from training microbiome data sets based on maximum likelihood. The posterior probability of a microbiome sample belonging to a disease or healthy category is calculated based on Bayes' theorem, using the likelihood values computed from the estimated Dirichlet-multinomial distribution, as well as a prior probability estimated from the training microbiome data set or previously published information on disease prevalence. When tested on real-world microbiome data sets, our method, called DMBC (for Dirichlet-multinomial Bayes classifier), shows better classification accuracy than the only existing Bayesian microbiome classifier based on a Dirichlet-multinomial mixture model and the popular random forest method. The advantage of DMBC is its built-in automatic feature selection, capable of identifying a subset of microbial taxa with the best classification accuracy between different classes of samples based on cross-validation. This unique ability enables DMBC to maintain and even improve its accuracy at modeling species-level taxa. The R package for DMBC is freely available at https://github.com/qunfengdong/DMBC. IMPORTANCE By incorporating prior information on disease prevalence, Bayes classifiers have the potential to estimate disease probability better than other common machine-learning methods. Thus, it is important to develop Bayes classifiers specifically tailored for microbiome data. Our method shows higher classification accuracy than the only existing Bayesian classifier and the popular random forest method, and thus provides an alternative option for using microbial compositions for disease diagnosis.

Land cover and land use mapping of the iSimangaliso Wetland Park, South Africa: comparison of oblique and orthogonal random forest algorithms

NASA Astrophysics Data System (ADS)

Bassa, Zaakirah; Bob, Urmilla; Szantoi, Zoltan; Ismail, Riyad

2016-01-01

In recent years, the popularity of tree-based ensemble methods for land cover classification has increased significantly. Using WorldView-2 image data, we evaluate the potential of the oblique random forest algorithm (oRF) to classify a highly heterogeneous protected area. In contrast to the random forest (RF) algorithm, the oRF algorithm builds multivariate trees by learning the optimal split using a supervised model. The oRF binary algorithm is adapted to a multiclass land cover and land use application using both the "one-against-one" and "one-against-all" combination approaches. Results show that the oRF algorithms are capable of achieving high classification accuracies (>80%). However, there was no statistical difference in classification accuracies obtained by the oRF algorithms and the more popular RF algorithm. For all the algorithms, user accuracies (UAs) and producer accuracies (PAs) >80% were recorded for most of the classes. Both the RF and oRF algorithms poorly classified the indigenous forest class as indicated by the low UAs and PAs. Finally, the results from this study advocate and support the utility of the oRF algorithm for land cover and land use mapping of protected areas using WorldView-2 image data.

Effect of sample size on multi-parametric prediction of tissue outcome in acute ischemic stroke using a random forest classifier

NASA Astrophysics Data System (ADS)

Forkert, Nils Daniel; Fiehler, Jens

2015-03-01

The tissue outcome prediction in acute ischemic stroke patients is highly relevant for clinical and research purposes. It has been shown that the combined analysis of diffusion and perfusion MRI datasets using high-level machine learning techniques leads to an improved prediction of final infarction compared to single perfusion parameter thresholding. However, most high-level classifiers require a previous training and, until now, it is ambiguous how many subjects are required for this, which is the focus of this work. 23 MRI datasets of acute stroke patients with known tissue outcome were used in this work. Relative values of diffusion and perfusion parameters as well as the binary tissue outcome were extracted on a voxel-by- voxel level for all patients and used for training of a random forest classifier. The number of patients used for training set definition was iteratively and randomly reduced from using all 22 other patients to only one other patient. Thus, 22 tissue outcome predictions were generated for each patient using the trained random forest classifiers and compared to the known tissue outcome using the Dice coefficient. Overall, a logarithmic relation between the number of patients used for training set definition and tissue outcome prediction accuracy was found. Quantitatively, a mean Dice coefficient of 0.45 was found for the prediction using the training set consisting of the voxel information from only one other patient, which increases to 0.53 if using all other patients (n=22). Based on extrapolation, 50-100 patients appear to be a reasonable tradeoff between tissue outcome prediction accuracy and effort required for data acquisition and preparation.

BitterSweetForest: A random forest based binary classifier to predict bitterness and sweetness of chemical compounds

NASA Astrophysics Data System (ADS)

Banerjee, Priyanka; Preissner, Robert

2018-04-01

Taste of a chemical compounds present in food stimulates us to take in nutrients and avoid poisons. However, the perception of taste greatly depends on the genetic as well as evolutionary perspectives. The aim of this work was the development and validation of a machine learning model based on molecular fingerprints to discriminate between sweet and bitter taste of molecules. BitterSweetForest is the first open access model based on KNIME workflow that provides platform for prediction of bitter and sweet taste of chemical compounds using molecular fingerprints and Random Forest based classifier. The constructed model yielded an accuracy of 95% and an AUC of 0.98 in cross-validation. In independent test set, BitterSweetForest achieved an accuracy of 96 % and an AUC of 0.98 for bitter and sweet taste prediction. The constructed model was further applied to predict the bitter and sweet taste of natural compounds, approved drugs as well as on an acute toxicity compound data set. BitterSweetForest suggests 70% of the natural product space, as bitter and 10 % of the natural product space as sweet with confidence score of 0.60 and above. 77 % of the approved drug set was predicted as bitter and 2% as sweet with a confidence scores of 0.75 and above. Similarly, 75% of the total compounds from acute oral toxicity class were predicted only as bitter with a minimum confidence score of 0.75, revealing toxic compounds are mostly bitter. Furthermore, we applied a Bayesian based feature analysis method to discriminate the most occurring chemical features between sweet and bitter compounds from the feature space of a circular fingerprint.

BitterSweetForest: A Random Forest Based Binary Classifier to Predict Bitterness and Sweetness of Chemical Compounds

PubMed Central

Banerjee, Priyanka; Preissner, Robert

2018-01-01

Taste of a chemical compound present in food stimulates us to take in nutrients and avoid poisons. However, the perception of taste greatly depends on the genetic as well as evolutionary perspectives. The aim of this work was the development and validation of a machine learning model based on molecular fingerprints to discriminate between sweet and bitter taste of molecules. BitterSweetForest is the first open access model based on KNIME workflow that provides platform for prediction of bitter and sweet taste of chemical compounds using molecular fingerprints and Random Forest based classifier. The constructed model yielded an accuracy of 95% and an AUC of 0.98 in cross-validation. In independent test set, BitterSweetForest achieved an accuracy of 96% and an AUC of 0.98 for bitter and sweet taste prediction. The constructed model was further applied to predict the bitter and sweet taste of natural compounds, approved drugs as well as on an acute toxicity compound data set. BitterSweetForest suggests 70% of the natural product space, as bitter and 10% of the natural product space as sweet with confidence score of 0.60 and above. 77% of the approved drug set was predicted as bitter and 2% as sweet with a confidence score of 0.75 and above. Similarly, 75% of the total compounds from acute oral toxicity class were predicted only as bitter with a minimum confidence score of 0.75, revealing toxic compounds are mostly bitter. Furthermore, we applied a Bayesian based feature analysis method to discriminate the most occurring chemical features between sweet and bitter compounds using the feature space of a circular fingerprint. PMID:29696137

Classifier for gravitational-wave inspiral signals in nonideal single-detector data

NASA Astrophysics Data System (ADS)

Kapadia, S. J.; Dent, T.; Dal Canton, T.

2017-11-01

We describe a multivariate classifier for candidate events in a templated search for gravitational-wave (GW) inspiral signals from neutron-star-black-hole (NS-BH) binaries, in data from ground-based detectors where sensitivity is limited by non-Gaussian noise transients. The standard signal-to-noise ratio (SNR) and chi-squared test for inspiral searches use only properties of a single matched filter at the time of an event; instead, we propose a classifier using features derived from a bank of inspiral templates around the time of each event, and also from a search using approximate sine-Gaussian templates. The classifier thus extracts additional information from strain data to discriminate inspiral signals from noise transients. We evaluate a random forest classifier on a set of single-detector events obtained from realistic simulated advanced LIGO data, using simulated NS-BH signals added to the data. The new classifier detects a factor of 1.5-2 more signals at low false positive rates as compared to the standard "reweighted SNR" statistic, and does not require the chi-squared test to be computed. Conversely, if only the SNR and chi-squared values of single-detector events are available, random forest classification performs nearly identically to the reweighted SNR.

Applying under-sampling techniques and cost-sensitive learning methods on risk assessment of breast cancer.

PubMed

Hsu, Jia-Lien; Hung, Ping-Cheng; Lin, Hung-Yen; Hsieh, Chung-Ho

2015-04-01

Breast cancer is one of the most common cause of cancer mortality. Early detection through mammography screening could significantly reduce mortality from breast cancer. However, most of screening methods may consume large amount of resources. We propose a computational model, which is solely based on personal health information, for breast cancer risk assessment. Our model can be served as a pre-screening program in the low-cost setting. In our study, the data set, consisting of 3976 records, is collected from Taipei City Hospital starting from 2008.1.1 to 2008.12.31. Based on the dataset, we first apply the sampling techniques and dimension reduction method to preprocess the testing data. Then, we construct various kinds of classifiers (including basic classifiers, ensemble methods, and cost-sensitive methods) to predict the risk. The cost-sensitive method with random forest classifier is able to achieve recall (or sensitivity) as 100 %. At the recall of 100 %, the precision (positive predictive value, PPV), and specificity of cost-sensitive method with random forest classifier was 2.9 % and 14.87 %, respectively. In our study, we build a breast cancer risk assessment model by using the data mining techniques. Our model has the potential to be served as an assisting tool in the breast cancer screening.

Automatic medical image annotation and keyword-based image retrieval using relevance feedback.

PubMed

Ko, Byoung Chul; Lee, JiHyeon; Nam, Jae-Yeal

2012-08-01

This paper presents novel multiple keywords annotation for medical images, keyword-based medical image retrieval, and relevance feedback method for image retrieval for enhancing image retrieval performance. For semantic keyword annotation, this study proposes a novel medical image classification method combining local wavelet-based center symmetric-local binary patterns with random forests. For keyword-based image retrieval, our retrieval system use the confidence score that is assigned to each annotated keyword by combining probabilities of random forests with predefined body relation graph. To overcome the limitation of keyword-based image retrieval, we combine our image retrieval system with relevance feedback mechanism based on visual feature and pattern classifier. Compared with other annotation and relevance feedback algorithms, the proposed method shows both improved annotation performance and accurate retrieval results.

CURE-SMOTE algorithm and hybrid algorithm for feature selection and parameter optimization based on random forests.

PubMed

Ma, Li; Fan, Suohai

2017-03-14

The random forests algorithm is a type of classifier with prominent universality, a wide application range, and robustness for avoiding overfitting. But there are still some drawbacks to random forests. Therefore, to improve the performance of random forests, this paper seeks to improve imbalanced data processing, feature selection and parameter optimization. We propose the CURE-SMOTE algorithm for the imbalanced data classification problem. Experiments on imbalanced UCI data reveal that the combination of Clustering Using Representatives (CURE) enhances the original synthetic minority oversampling technique (SMOTE) algorithms effectively compared with the classification results on the original data using random sampling, Borderline-SMOTE1, safe-level SMOTE, C-SMOTE, and k-means-SMOTE. Additionally, the hybrid RF (random forests) algorithm has been proposed for feature selection and parameter optimization, which uses the minimum out of bag (OOB) data error as its objective function. Simulation results on binary and higher-dimensional data indicate that the proposed hybrid RF algorithms, hybrid genetic-random forests algorithm, hybrid particle swarm-random forests algorithm and hybrid fish swarm-random forests algorithm can achieve the minimum OOB error and show the best generalization ability. The training set produced from the proposed CURE-SMOTE algorithm is closer to the original data distribution because it contains minimal noise. Thus, better classification results are produced from this feasible and effective algorithm. Moreover, the hybrid algorithm's F-value, G-mean, AUC and OOB scores demonstrate that they surpass the performance of the original RF algorithm. Hence, this hybrid algorithm provides a new way to perform feature selection and parameter optimization.

Random-Forest Classification of High-Resolution Remote Sensing Images and Ndsm Over Urban Areas

NASA Astrophysics Data System (ADS)

Sun, X. F.; Lin, X. G.

2017-09-01

As an intermediate step between raw remote sensing data and digital urban maps, remote sensing data classification has been a challenging and long-standing research problem in the community of remote sensing. In this work, an effective classification method is proposed for classifying high-resolution remote sensing data over urban areas. Starting from high resolution multi-spectral images and 3D geometry data, our method proceeds in three main stages: feature extraction, classification, and classified result refinement. First, we extract color, vegetation index and texture features from the multi-spectral image and compute the height, elevation texture and differential morphological profile (DMP) features from the 3D geometry data. Then in the classification stage, multiple random forest (RF) classifiers are trained separately, then combined to form a RF ensemble to estimate each sample's category probabilities. Finally the probabilities along with the feature importance indicator outputted by RF ensemble are used to construct a fully connected conditional random field (FCCRF) graph model, by which the classification results are refined through mean-field based statistical inference. Experiments on the ISPRS Semantic Labeling Contest dataset show that our proposed 3-stage method achieves 86.9% overall accuracy on the test data.

Differentiation of fat, muscle, and edema in thigh MRIs using random forest classification

NASA Astrophysics Data System (ADS)

Kovacs, William; Liu, Chia-Ying; Summers, Ronald M.; Yao, Jianhua

2016-03-01

There are many diseases that affect the distribution of muscles, including Duchenne and fascioscapulohumeral dystrophy among other myopathies. In these disease cases, it is important to quantify both the muscle and fat volumes to track the disease progression. There has also been evidence that abnormal signal intensity on the MR images, which often is an indication of edema or inflammation can be a good predictor for muscle deterioration. We present a fully-automated method that examines magnetic resonance (MR) images of the thigh and identifies the fat, muscle, and edema using a random forest classifier. First the thigh regions are automatically segmented using the T1 sequence. Then, inhomogeneity artifacts were corrected using the N3 technique. The T1 and STIR (short tau inverse recovery) images are then aligned using landmark based registration with the bone marrow. The normalized T1 and STIR intensity values are used to train the random forest. Once trained, the random forest can accurately classify the aforementioned classes. This method was evaluated on MR images of 9 patients. The precision values are 0.91+/-0.06, 0.98+/-0.01 and 0.50+/-0.29 for muscle, fat, and edema, respectively. The recall values are 0.95+/-0.02, 0.96+/-0.03 and 0.43+/-0.09 for muscle, fat, and edema, respectively. This demonstrates the feasibility of utilizing information from multiple MR sequences for the accurate quantification of fat, muscle and edema.

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.

Evaluation of Semi-supervised Learning for Classification of Protein Crystallization Imagery

PubMed Central

Sigdel, Madhav; Dinç, İmren; Dinç, Semih; Sigdel, Madhu S.; Pusey, Marc L.; Aygün, Ramazan S.

2015-01-01

In this paper, we investigate the performance of two wrapper methods for semi-supervised learning algorithms for classification of protein crystallization images with limited labeled images. Firstly, we evaluate the performance of semi-supervised approach using self-training with naïve Bayesian (NB) and sequential minimum optimization (SMO) as the base classifiers. The confidence values returned by these classifiers are used to select high confident predictions to be used for self-training. Secondly, we analyze the performance of Yet Another Two Stage Idea (YATSI) semi-supervised learning using NB, SMO, multilayer perceptron (MLP), J48 and random forest (RF) classifiers. These results are compared with the basic supervised learning using the same training sets. We perform our experiments on a dataset consisting of 2250 protein crystallization images for different proportions of training and test data. Our results indicate that NB and SMO using both self-training and YATSI semi-supervised approaches improve accuracies with respect to supervised learning. On the other hand, MLP, J48 and RF perform better using basic supervised learning. Overall, random forest classifier yields the best accuracy with supervised learning for our dataset. PMID:25914518

Analyzing Body Movements within the Laban Effort Framework Using a Single Accelerometer

PubMed Central

Kikhia, Basel; Gomez, Miguel; Jiménez, Lara Lorna; Hallberg, Josef; Karvonen, Niklas; Synnes, Kåre

2014-01-01

This article presents a study on analyzing body movements by using a single accelerometer sensor. The investigated categories of body movements belong to the Laban Effort Framework: Strong—Light, Free—Bound and Sudden—Sustained. All body movements were represented by a set of activities used for data collection. The calculated accuracy of detecting the body movements was based on collecting data from a single wireless tri-axial accelerometer sensor. Ten healthy subjects collected data from three body locations (chest, wrist and thigh) simultaneously in order to analyze the locations comparatively. The data was then processed and analyzed using Machine Learning techniques. The wrist placement was found to be the best single location to record data for detecting Strong—Light body movements using the Random Forest classifier. The wrist placement was also the best location for classifying Bound—Free body movements using the SVM classifier. However, the data collected from the chest placement yielded the best results for detecting Sudden—Sustained body movements using the Random Forest classifier. The study shows that the choice of the accelerometer placement should depend on the targeted type of movement. In addition, the choice of the classifier when processing data should also depend on the chosen location and the target movement. PMID:24662408

Beyond where to how: a machine learning approach for sensing mobility contexts using smartphone sensors.

PubMed

Guinness, Robert E

2015-04-28

This paper presents the results of research on the use of smartphone sensors (namely, GPS and accelerometers), geospatial information (points of interest, such as bus stops and train stations) and machine learning (ML) to sense mobility contexts. Our goal is to develop techniques to continuously and automatically detect a smartphone user's mobility activities, including walking, running, driving and using a bus or train, in real-time or near-real-time (<5 s). We investigated a wide range of supervised learning techniques for classification, including decision trees (DT), support vector machines (SVM), naive Bayes classifiers (NB), Bayesian networks (BN), logistic regression (LR), artificial neural networks (ANN) and several instance-based classifiers (KStar, LWLand IBk). Applying ten-fold cross-validation, the best performers in terms of correct classification rate (i.e., recall) were DT (96.5%), BN (90.9%), LWL (95.5%) and KStar (95.6%). In particular, the DT-algorithm RandomForest exhibited the best overall performance. After a feature selection process for a subset of algorithms, the performance was improved slightly. Furthermore, after tuning the parameters of RandomForest, performance improved to above 97.5%. Lastly, we measured the computational complexity of the classifiers, in terms of central processing unit (CPU) time needed for classification, to provide a rough comparison between the algorithms in terms of battery usage requirements. As a result, the classifiers can be ranked from lowest to highest complexity (i.e., computational cost) as follows: SVM, ANN, LR, BN, DT, NB, IBk, LWL and KStar. The instance-based classifiers take considerably more computational time than the non-instance-based classifiers, whereas the slowest non-instance-based classifier (NB) required about five-times the amount of CPU time as the fastest classifier (SVM). The above results suggest that DT algorithms are excellent candidates for detecting mobility contexts in smartphones, both in terms of performance and computational complexity.

Beyond Where to How: A Machine Learning Approach for Sensing Mobility Contexts Using Smartphone Sensors †

PubMed Central

Guinness, Robert E.

2015-01-01

This paper presents the results of research on the use of smartphone sensors (namely, GPS and accelerometers), geospatial information (points of interest, such as bus stops and train stations) and machine learning (ML) to sense mobility contexts. Our goal is to develop techniques to continuously and automatically detect a smartphone user's mobility activities, including walking, running, driving and using a bus or train, in real-time or near-real-time (<5 s). We investigated a wide range of supervised learning techniques for classification, including decision trees (DT), support vector machines (SVM), naive Bayes classifiers (NB), Bayesian networks (BN), logistic regression (LR), artificial neural networks (ANN) and several instance-based classifiers (KStar, LWLand IBk). Applying ten-fold cross-validation, the best performers in terms of correct classification rate (i.e., recall) were DT (96.5%), BN (90.9%), LWL (95.5%) and KStar (95.6%). In particular, the DT-algorithm RandomForest exhibited the best overall performance. After a feature selection process for a subset of algorithms, the performance was improved slightly. Furthermore, after tuning the parameters of RandomForest, performance improved to above 97.5%. Lastly, we measured the computational complexity of the classifiers, in terms of central processing unit (CPU) time needed for classification, to provide a rough comparison between the algorithms in terms of battery usage requirements. As a result, the classifiers can be ranked from lowest to highest complexity (i.e., computational cost) as follows: SVM, ANN, LR, BN, DT, NB, IBk, LWL and KStar. The instance-based classifiers take considerably more computational time than the non-instance-based classifiers, whereas the slowest non-instance-based classifier (NB) required about five-times the amount of CPU time as the fastest classifier (SVM). The above results suggest that DT algorithms are excellent candidates for detecting mobility contexts in smartphones, both in terms of performance and computational complexity. PMID:25928060

Modeling Verdict Outcomes Using Social Network Measures: The Watergate and Caviar Network Cases.

PubMed

Masías, Víctor Hugo; Valle, Mauricio; Morselli, Carlo; Crespo, Fernando; Vargas, Augusto; Laengle, Sigifredo

2016-01-01

Modelling criminal trial verdict outcomes using social network measures is an emerging research area in quantitative criminology. Few studies have yet analyzed which of these measures are the most important for verdict modelling or which data classification techniques perform best for this application. To compare the performance of different techniques in classifying members of a criminal network, this article applies three different machine learning classifiers-Logistic Regression, Naïve Bayes and Random Forest-with a range of social network measures and the necessary databases to model the verdicts in two real-world cases: the U.S. Watergate Conspiracy of the 1970's and the now-defunct Canada-based international drug trafficking ring known as the Caviar Network. In both cases it was found that the Random Forest classifier did better than either Logistic Regression or Naïve Bayes, and its superior performance was statistically significant. This being so, Random Forest was used not only for classification but also to assess the importance of the measures. For the Watergate case, the most important one proved to be betweenness centrality while for the Caviar Network, it was the effective size of the network. These results are significant because they show that an approach combining machine learning with social network analysis not only can generate accurate classification models but also helps quantify the importance social network variables in modelling verdict outcomes. We conclude our analysis with a discussion and some suggestions for future work in verdict modelling using social network measures.

Random forest classification of stars in the Galactic Centre

NASA Astrophysics Data System (ADS)

Plewa, P. M.

2018-05-01

Near-infrared high-angular resolution imaging observations of the Milky Way's nuclear star cluster have revealed all luminous members of the existing stellar population within the central parsec. Generally, these stars are either evolved late-type giants or massive young, early-type stars. We revisit the problem of stellar classification based on intermediate-band photometry in the K band, with the primary aim of identifying faint early-type candidate stars in the extended vicinity of the central massive black hole. A random forest classifier, trained on a subsample of spectroscopically identified stars, performs similarly well as competitive methods (F1 = 0.85), without involving any model of stellar spectral energy distributions. Advantages of using such a machine-trained classifier are a minimum of required calibration effort, a predictive accuracy expected to improve as more training data become available, and the ease of application to future, larger data sets. By applying this classifier to archive data, we are also able to reproduce the results of previous studies of the spatial distribution and the K-band luminosity function of both the early- and late-type stars.

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.

Combination of support vector machine, artificial neural network and random forest for improving the classification of convective and stratiform rain using spectral features of SEVIRI data

NASA Astrophysics Data System (ADS)

Lazri, Mourad; Ameur, Soltane

2018-05-01

A model combining three classifiers, namely Support vector machine, Artificial neural network and Random forest (SAR) is designed for improving the classification of convective and stratiform rain. This model (SAR model) has been trained and then tested on a datasets derived from MSG-SEVIRI (Meteosat Second Generation-Spinning Enhanced Visible and Infrared Imager). Well-classified, mid-classified and misclassified pixels are determined from the combination of three classifiers. Mid-classified and misclassified pixels that are considered unreliable pixels are reclassified by using a novel training of the developed scheme. In this novel training, only the input data corresponding to the pixels in question to are used. This whole process is repeated a second time and applied to mid-classified and misclassified pixels separately. Learning and validation of the developed scheme are realized against co-located data observed by ground radar. The developed scheme outperformed different classifiers used separately and reached 97.40% of overall accuracy of classification.

Latent information in fluency lists predicts functional decline in persons at risk for Alzheimer disease.

PubMed

Clark, D G; Kapur, P; Geldmacher, D S; Brockington, J C; Harrell, L; DeRamus, T P; Blanton, P D; Lokken, K; Nicholas, A P; Marson, D C

2014-06-01

We constructed random forest classifiers employing either the traditional method of scoring semantic fluency word lists or new methods. These classifiers were then compared in terms of their ability to diagnose Alzheimer disease (AD) or to prognosticate among individuals along the continuum from cognitively normal (CN) through mild cognitive impairment (MCI) to AD. Semantic fluency lists from 44 cognitively normal elderly individuals, 80 MCI patients, and 41 AD patients were transcribed into electronic text files and scored by four methods: traditional raw scores, clustering and switching scores, "generalized" versions of clustering and switching, and a method based on independent components analysis (ICA). Random forest classifiers based on raw scores were compared to "augmented" classifiers that incorporated newer scoring methods. Outcome variables included AD diagnosis at baseline, MCI conversion, increase in Clinical Dementia Rating-Sum of Boxes (CDR-SOB) score, or decrease in Financial Capacity Instrument (FCI) score. Receiver operating characteristic (ROC) curves were constructed for each classifier and the area under the curve (AUC) was calculated. We compared AUC between raw and augmented classifiers using Delong's test and assessed validity and reliability of the augmented classifier. Augmented classifiers outperformed classifiers based on raw scores for the outcome measures AD diagnosis (AUC .97 vs. .95), MCI conversion (AUC .91 vs. .77), CDR-SOB increase (AUC .90 vs. .79), and FCI decrease (AUC .89 vs. .72). Measures of validity and stability over time support the use of the method. Latent information in semantic fluency word lists is useful for predicting cognitive and functional decline among elderly individuals at increased risk for developing AD. Modern machine learning methods may incorporate latent information to enhance the diagnostic value of semantic fluency raw scores. These methods could yield information valuable for patient care and clinical trial design with a relatively small investment of time and money. Published by Elsevier Ltd.

Multi-label spacecraft electrical signal classification method based on DBN and random forest

PubMed Central

Li, Ke; Yu, Nan; Li, Pengfei; Song, Shimin; Wu, Yalei; Li, Yang; Liu, Meng

2017-01-01

In spacecraft electrical signal characteristic data, there exists a large amount of data with high-dimensional features, a high computational complexity degree, and a low rate of identification problems, which causes great difficulty in fault diagnosis of spacecraft electronic load systems. This paper proposes a feature extraction method that is based on deep belief networks (DBN) and a classification method that is based on the random forest (RF) algorithm; The proposed algorithm mainly employs a multi-layer neural network to reduce the dimension of the original data, and then, classification is applied. Firstly, we use the method of wavelet denoising, which was used to pre-process the data. Secondly, the deep belief network is used to reduce the feature dimension and improve the rate of classification for the electrical characteristics data. Finally, we used the random forest algorithm to classify the data and comparing it with other algorithms. The experimental results show that compared with other algorithms, the proposed method shows excellent performance in terms of accuracy, computational efficiency, and stability in addressing spacecraft electrical signal data. PMID:28486479

Multi-label spacecraft electrical signal classification method based on DBN and random forest.

PubMed

Li, Ke; Yu, Nan; Li, Pengfei; Song, Shimin; Wu, Yalei; Li, Yang; Liu, Meng

2017-01-01

In spacecraft electrical signal characteristic data, there exists a large amount of data with high-dimensional features, a high computational complexity degree, and a low rate of identification problems, which causes great difficulty in fault diagnosis of spacecraft electronic load systems. This paper proposes a feature extraction method that is based on deep belief networks (DBN) and a classification method that is based on the random forest (RF) algorithm; The proposed algorithm mainly employs a multi-layer neural network to reduce the dimension of the original data, and then, classification is applied. Firstly, we use the method of wavelet denoising, which was used to pre-process the data. Secondly, the deep belief network is used to reduce the feature dimension and improve the rate of classification for the electrical characteristics data. Finally, we used the random forest algorithm to classify the data and comparing it with other algorithms. The experimental results show that compared with other algorithms, the proposed method shows excellent performance in terms of accuracy, computational efficiency, and stability in addressing spacecraft electrical signal data.

Tree species classification in subtropical forests using small-footprint full-waveform LiDAR data

NASA Astrophysics Data System (ADS)

Cao, Lin; Coops, Nicholas C.; Innes, John L.; Dai, Jinsong; Ruan, Honghua; She, Guanghui

2016-07-01

The accurate classification of tree species is critical for the management of forest ecosystems, particularly subtropical forests, which are highly diverse and complex ecosystems. While airborne Light Detection and Ranging (LiDAR) technology offers significant potential to estimate forest structural attributes, the capacity of this new tool to classify species is less well known. In this research, full-waveform metrics were extracted by a voxel-based composite waveform approach and examined with a Random Forests classifier to discriminate six subtropical tree species (i.e., Masson pine (Pinus massoniana Lamb.)), Chinese fir (Cunninghamia lanceolata (Lamb.) Hook.), Slash pines (Pinus elliottii Engelm.), Sawtooth oak (Quercus acutissima Carruth.) and Chinese holly (Ilex chinensis Sims.) at three levels of discrimination. As part of the analysis, the optimal voxel size for modelling the composite waveforms was investigated, the most important predictor metrics for species classification assessed and the effect of scan angle on species discrimination examined. Results demonstrate that all tree species were classified with relatively high accuracy (68.6% for six classes, 75.8% for four main species and 86.2% for conifers and broadleaved trees). Full-waveform metrics (based on height of median energy, waveform distance and number of waveform peaks) demonstrated high classification importance and were stable among various voxel sizes. The results also suggest that the voxel based approach can alleviate some of the issues associated with large scan angles. In summary, the results indicate that full-waveform LIDAR data have significant potential for tree species classification in the subtropical forests.

Exploring diversity in ensemble classification: Applications in large area land cover mapping

NASA Astrophysics Data System (ADS)

Mellor, Andrew; Boukir, Samia

2017-07-01

Ensemble classifiers, such as random forests, are now commonly applied in the field of remote sensing, and have been shown to perform better than single classifier systems, resulting in reduced generalisation error. Diversity across the members of ensemble classifiers is known to have a strong influence on classification performance - whereby classifier errors are uncorrelated and more uniformly distributed across ensemble members. The relationship between ensemble diversity and classification performance has not yet been fully explored in the fields of information science and machine learning and has never been examined in the field of remote sensing. This study is a novel exploration of ensemble diversity and its link to classification performance, applied to a multi-class canopy cover classification problem using random forests and multisource remote sensing and ancillary GIS data, across seven million hectares of diverse dry-sclerophyll dominated public forests in Victoria Australia. A particular emphasis is placed on analysing the relationship between ensemble diversity and ensemble margin - two key concepts in ensemble learning. The main novelty of our work is on boosting diversity by emphasizing the contribution of lower margin instances used in the learning process. Exploring the influence of tree pruning on diversity is also a new empirical analysis that contributes to a better understanding of ensemble performance. Results reveal insights into the trade-off between ensemble classification accuracy and diversity, and through the ensemble margin, demonstrate how inducing diversity by targeting lower margin training samples is a means of achieving better classifier performance for more difficult or rarer classes and reducing information redundancy in classification problems. Our findings inform strategies for collecting training data and designing and parameterising ensemble classifiers, such as random forests. This is particularly important in large area remote sensing applications, for which training data is costly and resource intensive to collect.

Classification of coronary artery tissues using optical coherence tomography imaging in Kawasaki disease

NASA Astrophysics Data System (ADS)

Abdolmanafi, Atefeh; Prasad, Arpan Suravi; Duong, Luc; Dahdah, Nagib

2016-03-01

Intravascular imaging modalities, such as Optical Coherence Tomography (OCT) allow nowadays improving diagnosis, treatment, follow-up, and even prevention of coronary artery disease in the adult. OCT has been recently used in children following Kawasaki disease (KD), the most prevalent acquired coronary artery disease during childhood with devastating complications. The assessment of coronary artery layers with OCT and early detection of coronary sequelae secondary to KD is a promising tool for preventing myocardial infarction in this population. More importantly, OCT is promising for tissue quantification of the inner vessel wall, including neo intima luminal myofibroblast proliferation, calcification, and fibrous scar deposits. The goal of this study is to classify the coronary artery layers of OCT imaging obtained from a series of KD patients. Our approach is focused on developing a robust Random Forest classifier built on the idea of randomly selecting a subset of features at each node and based on second- and higher-order statistical texture analysis which estimates the gray-level spatial distribution of images by specifying the local features of each pixel and extracting the statistics from their distribution. The average classification accuracy for intima and media are 76.36% and 73.72% respectively. Random forest classifier with texture analysis promises for classification of coronary artery tissue.

Accurate Diabetes Risk Stratification Using Machine Learning: Role of Missing Value and Outliers.

PubMed

Maniruzzaman, Md; Rahman, Md Jahanur; Al-MehediHasan, Md; Suri, Harman S; Abedin, Md Menhazul; El-Baz, Ayman; Suri, Jasjit S

2018-04-10

Diabetes mellitus is a group of metabolic diseases in which blood sugar levels are too high. About 8.8% of the world was diabetic in 2017. It is projected that this will reach nearly 10% by 2045. The major challenge is that when machine learning-based classifiers are applied to such data sets for risk stratification, leads to lower performance. Thus, our objective is to develop an optimized and robust machine learning (ML) system under the assumption that missing values or outliers if replaced by a median configuration will yield higher risk stratification accuracy. This ML-based risk stratification is designed, optimized and evaluated, where: (i) the features are extracted and optimized from the six feature selection techniques (random forest, logistic regression, mutual information, principal component analysis, analysis of variance, and Fisher discriminant ratio) and combined with ten different types of classifiers (linear discriminant analysis, quadratic discriminant analysis, naïve Bayes, Gaussian process classification, support vector machine, artificial neural network, Adaboost, logistic regression, decision tree, and random forest) under the hypothesis that both missing values and outliers when replaced by computed medians will improve the risk stratification accuracy. Pima Indian diabetic dataset (768 patients: 268 diabetic and 500 controls) was used. Our results demonstrate that on replacing the missing values and outliers by group median and median values, respectively and further using the combination of random forest feature selection and random forest classification technique yields an accuracy, sensitivity, specificity, positive predictive value, negative predictive value and area under the curve as: 92.26%, 95.96%, 79.72%, 91.14%, 91.20%, and 0.93, respectively. This is an improvement of 10% over previously developed techniques published in literature. The system was validated for its stability and reliability. RF-based model showed the best performance when outliers are replaced by median values.

A k-mer-based barcode DNA classification methodology based on spectral representation and a neural gas network.

PubMed

Fiannaca, Antonino; La Rosa, Massimo; Rizzo, Riccardo; Urso, Alfonso

2015-07-01

In this paper, an alignment-free method for DNA barcode classification that is based on both a spectral representation and a neural gas network for unsupervised clustering is proposed. In the proposed methodology, distinctive words are identified from a spectral representation of DNA sequences. A taxonomic classification of the DNA sequence is then performed using the sequence signature, i.e., the smallest set of k-mers that can assign a DNA sequence to its proper taxonomic category. Experiments were then performed to compare our method with other supervised machine learning classification algorithms, such as support vector machine, random forest, ripper, naïve Bayes, ridor, and classification tree, which also consider short DNA sequence fragments of 200 and 300 base pairs (bp). The experimental tests were conducted over 10 real barcode datasets belonging to different animal species, which were provided by the on-line resource "Barcode of Life Database". The experimental results showed that our k-mer-based approach is directly comparable, in terms of accuracy, recall and precision metrics, with the other classifiers when considering full-length sequences. In addition, we demonstrate the robustness of our method when a classification is performed task with a set of short DNA sequences that were randomly extracted from the original data. For example, the proposed method can reach the accuracy of 64.8% at the species level with 200-bp fragments. Under the same conditions, the best other classifier (random forest) reaches the accuracy of 20.9%. Our results indicate that we obtained a clear improvement over the other classifiers for the study of short DNA barcode sequence fragments. Copyright © 2015 Elsevier B.V. All rights reserved.

Random Bits Forest: a Strong Classifier/Regressor for Big Data

NASA Astrophysics Data System (ADS)

Wang, Yi; Li, Yi; Pu, Weilin; Wen, Kathryn; Shugart, Yin Yao; Xiong, Momiao; Jin, Li

2016-07-01

Efficiency, memory consumption, and robustness are common problems with many popular methods for data analysis. As a solution, we present Random Bits Forest (RBF), a classification and regression algorithm that integrates neural networks (for depth), boosting (for width), and random forests (for prediction accuracy). Through a gradient boosting scheme, it first generates and selects ~10,000 small, 3-layer random neural networks. These networks are then fed into a modified random forest algorithm to obtain predictions. Testing with datasets from the UCI (University of California, Irvine) Machine Learning Repository shows that RBF outperforms other popular methods in both accuracy and robustness, especially with large datasets (N > 1000). The algorithm also performed highly in testing with an independent data set, a real psoriasis genome-wide association study (GWAS).

GPURFSCREEN: a GPU based virtual screening tool using random forest classifier.

PubMed

Jayaraj, P B; Ajay, Mathias K; Nufail, M; Gopakumar, G; Jaleel, U C A

2016-01-01

In-silico methods are an integral part of modern drug discovery paradigm. Virtual screening, an in-silico method, is used to refine data models and reduce the chemical space on which wet lab experiments need to be performed. Virtual screening of a ligand data model requires large scale computations, making it a highly time consuming task. This process can be speeded up by implementing parallelized algorithms on a Graphical Processing Unit (GPU). Random Forest is a robust classification algorithm that can be employed in the virtual screening. A ligand based virtual screening tool (GPURFSCREEN) that uses random forests on GPU systems has been proposed and evaluated in this paper. This tool produces optimized results at a lower execution time for large bioassay data sets. The quality of results produced by our tool on GPU is same as that on a regular serial environment. Considering the magnitude of data to be screened, the parallelized virtual screening has a significantly lower running time at high throughput. The proposed parallel tool outperforms its serial counterpart by successfully screening billions of molecules in training and prediction phases.

Source identification of western Oregon Douglas-fir wood cores using mass spectrometry and random forest classification.

PubMed

Finch, Kristen; Espinoza, Edgard; Jones, F Andrew; Cronn, Richard

2017-05-01

We investigated whether wood metabolite profiles from direct analysis in real time (time-of-flight) mass spectrometry (DART-TOFMS) could be used to determine the geographic origin of Douglas-fir wood cores originating from two regions in western Oregon, USA. Three annual ring mass spectra were obtained from 188 adult Douglas-fir trees, and these were analyzed using random forest models to determine whether samples could be classified to geographic origin, growth year, or growth year and geographic origin. Specific wood molecules that contributed to geographic discrimination were identified. Douglas-fir mass spectra could be differentiated into two geographic classes with an accuracy between 70% and 76%. Classification models could not accurately classify sample mass spectra based on growth year. Thirty-two molecules were identified as key for classifying western Oregon Douglas-fir wood cores to geographic origin. DART-TOFMS is capable of detecting minute but regionally informative differences in wood molecules over a small geographic scale, and these differences made it possible to predict the geographic origin of Douglas-fir wood with moderate accuracy. Studies involving DART-TOFMS, alone and in combination with other technologies, will be relevant for identifying the geographic origin of illegally harvested wood.

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

PPCM: Combing multiple classifiers to improve protein-protein interaction prediction

DOE PAGES

Yao, Jianzhuang; Guo, Hong; Yang, Xiaohan

2015-08-01

Determining protein-protein interaction (PPI) in biological systems is of considerable importance, and prediction of PPI has become a popular research area. Although different classifiers have been developed for PPI prediction, no single classifier seems to be able to predict PPI with high confidence. We postulated that by combining individual classifiers the accuracy of PPI prediction could be improved. We developed a method called protein-protein interaction prediction classifiers merger (PPCM), and this method combines output from two PPI prediction tools, GO2PPI and Phyloprof, using Random Forests algorithm. The performance of PPCM was tested by area under the curve (AUC) using anmore » assembled Gold Standard database that contains both positive and negative PPI pairs. Our AUC test showed that PPCM significantly improved the PPI prediction accuracy over the corresponding individual classifiers. We found that additional classifiers incorporated into PPCM could lead to further improvement in the PPI prediction accuracy. Furthermore, cross species PPCM could achieve competitive and even better prediction accuracy compared to the single species PPCM. This study established a robust pipeline for PPI prediction by integrating multiple classifiers using Random Forests algorithm. Ultimately, this pipeline will be useful for predicting PPI in nonmodel species.« less

Large unbalanced credit scoring using Lasso-logistic regression ensemble.

PubMed

Wang, Hong; Xu, Qingsong; Zhou, Lifeng

2015-01-01

Recently, various ensemble learning methods with different base classifiers have been proposed for credit scoring problems. However, for various reasons, there has been little research using logistic regression as the base classifier. In this paper, given large unbalanced data, we consider the plausibility of ensemble learning using regularized logistic regression as the base classifier to deal with credit scoring problems. In this research, the data is first balanced and diversified by clustering and bagging algorithms. Then we apply a Lasso-logistic regression learning ensemble to evaluate the credit risks. We show that the proposed algorithm outperforms popular credit scoring models such as decision tree, Lasso-logistic regression and random forests in terms of AUC and F-measure. We also provide two importance measures for the proposed model to identify important variables in the data.

Comparison of classification methods for voxel-based prediction of acute ischemic stroke outcome following intra-arterial intervention

NASA Astrophysics Data System (ADS)

Winder, Anthony J.; Siemonsen, Susanne; Flottmann, Fabian; Fiehler, Jens; Forkert, Nils D.

2017-03-01

Voxel-based tissue outcome prediction in acute ischemic stroke patients is highly relevant for both clinical routine and research. Previous research has shown that features extracted from baseline multi-parametric MRI datasets have a high predictive value and can be used for the training of classifiers, which can generate tissue outcome predictions for both intravenous and conservative treatments. However, with the recent advent and popularization of intra-arterial thrombectomy treatment, novel research specifically addressing the utility of predictive classi- fiers for thrombectomy intervention is necessary for a holistic understanding of current stroke treatment options. The aim of this work was to develop three clinically viable tissue outcome prediction models using approximate nearest-neighbor, generalized linear model, and random decision forest approaches and to evaluate the accuracy of predicting tissue outcome after intra-arterial treatment. Therefore, the three machine learning models were trained, evaluated, and compared using datasets of 42 acute ischemic stroke patients treated with intra-arterial thrombectomy. Classifier training utilized eight voxel-based features extracted from baseline MRI datasets and five global features. Evaluation of classifier-based predictions was performed via comparison to the known tissue outcome, which was determined in follow-up imaging, using the Dice coefficient and leave-on-patient-out cross validation. The random decision forest prediction model led to the best tissue outcome predictions with a mean Dice coefficient of 0.37. The approximate nearest-neighbor and generalized linear model performed equally suboptimally with average Dice coefficients of 0.28 and 0.27 respectively, suggesting that both non-linearity and machine learning are desirable properties of a classifier well-suited to the intra-arterial tissue outcome prediction problem.

Identification of species based on DNA barcode using k-mer feature vector and Random forest classifier.

PubMed

Meher, Prabina Kumar; Sahu, Tanmaya Kumar; Rao, A R

2016-11-05

DNA barcoding is a molecular diagnostic method that allows automated and accurate identification of species based on a short and standardized fragment of DNA. To this end, an attempt has been made in this study to develop a computational approach for identifying the species by comparing its barcode with the barcode sequence of known species present in the reference library. Each barcode sequence was first mapped onto a numeric feature vector based on k-mer frequencies and then Random forest methodology was employed on the transformed dataset for species identification. The proposed approach outperformed similarity-based, tree-based, diagnostic-based approaches and found comparable with existing supervised learning based approaches in terms of species identification success rate, while compared using real and simulated datasets. Based on the proposed approach, an online web interface SPIDBAR has also been developed and made freely available at http://cabgrid.res.in:8080/spidbar/ for species identification by the taxonomists. Copyright © 2016 Elsevier B.V. All rights reserved.

Modification of the random forest algorithm to avoid statistical dependence problems when classifying remote sensing imagery

NASA Astrophysics Data System (ADS)

Cánovas-García, Fulgencio; Alonso-Sarría, Francisco; Gomariz-Castillo, Francisco; Oñate-Valdivieso, Fernando

2017-06-01

Random forest is a classification technique widely used in remote sensing. One of its advantages is that it produces an estimation of classification accuracy based on the so called out-of-bag cross-validation method. It is usually assumed that such estimation is not biased and may be used instead of validation based on an external data-set or a cross-validation external to the algorithm. In this paper we show that this is not necessarily the case when classifying remote sensing imagery using training areas with several pixels or objects. According to our results, out-of-bag cross-validation clearly overestimates accuracy, both overall and per class. The reason is that, in a training patch, pixels or objects are not independent (from a statistical point of view) of each other; however, they are split by bootstrapping into in-bag and out-of-bag as if they were really independent. We believe that putting whole patch, rather than pixels/objects, in one or the other set would produce a less biased out-of-bag cross-validation. To deal with the problem, we propose a modification of the random forest algorithm to split training patches instead of the pixels (or objects) that compose them. This modified algorithm does not overestimate accuracy and has no lower predictive capability than the original. When its results are validated with an external data-set, the accuracy is not different from that obtained with the original algorithm. We analysed three remote sensing images with different classification approaches (pixel and object based); in the three cases reported, the modification we propose produces a less biased accuracy estimation.

A random forest model based classification scheme for neonatal amplitude-integrated EEG.

PubMed

Chen, Weiting; Wang, Yu; Cao, Guitao; Chen, Guoqiang; Gu, Qiufang

2014-01-01

Modern medical advances have greatly increased the survival rate of infants, while they remain in the higher risk group for neurological problems later in life. For the infants with encephalopathy or seizures, identification of the extent of brain injury is clinically challenging. Continuous amplitude-integrated electroencephalography (aEEG) monitoring offers a possibility to directly monitor the brain functional state of the newborns over hours, and has seen an increasing application in neonatal intensive care units (NICUs). This paper presents a novel combined feature set of aEEG and applies random forest (RF) method to classify aEEG tracings. To that end, a series of experiments were conducted on 282 aEEG tracing cases (209 normal and 73 abnormal ones). Basic features, statistic features and segmentation features were extracted from both the tracing as a whole and the segmented recordings, and then form a combined feature set. All the features were sent to a classifier afterwards. The significance of feature, the data segmentation, the optimization of RF parameters, and the problem of imbalanced datasets were examined through experiments. Experiments were also done to evaluate the performance of RF on aEEG signal classifying, compared with several other widely used classifiers including SVM-Linear, SVM-RBF, ANN, Decision Tree (DT), Logistic Regression(LR), ML, and LDA. The combined feature set can better characterize aEEG signals, compared with basic features, statistic features and segmentation features respectively. With the combined feature set, the proposed RF-based aEEG classification system achieved a correct rate of 92.52% and a high F1-score of 95.26%. Among all of the seven classifiers examined in our work, the RF method got the highest correct rate, sensitivity, specificity, and F1-score, which means that RF outperforms all of the other classifiers considered here. The results show that the proposed RF-based aEEG classification system with the combined feature set is efficient and helpful to better detect the brain disorders in newborns.

Forest community classification of the Porcupine River drainage, interior Alaska, and its application to forest management.

Treesearch

John Yarie

1983-01-01

The forest vegetation of 3,600,000 hectares in northeast interior Alaska was classified. A total of 365 plots located in a stratified random design were run through the ordination programs SIMORD and TWINSPAN. A total of 40 forest communities were described vegetatively and, to a limited extent, environmentally. The area covered by each community was similar, ranging...

Pigmented skin lesion detection using random forest and wavelet-based texture

NASA Astrophysics Data System (ADS)

Hu, Ping; Yang, Tie-jun

2016-10-01

The incidence of cutaneous malignant melanoma, a disease of worldwide distribution and is the deadliest form of skin cancer, has been rapidly increasing over the last few decades. Because advanced cutaneous melanoma is still incurable, early detection is an important step toward a reduction in mortality. Dermoscopy photographs are commonly used in melanoma diagnosis and can capture detailed features of a lesion. A great variability exists in the visual appearance of pigmented skin lesions. Therefore, in order to minimize the diagnostic errors that result from the difficulty and subjectivity of visual interpretation, an automatic detection approach is required. The objectives of this paper were to propose a hybrid method using random forest and Gabor wavelet transformation to accurately differentiate which part belong to lesion area and the other is not in a dermoscopy photographs and analyze segmentation accuracy. A random forest classifier consisting of a set of decision trees was used for classification. Gabor wavelets transformation are the mathematical model of visual cortical cells of mammalian brain and an image can be decomposed into multiple scales and multiple orientations by using it. The Gabor function has been recognized as a very useful tool in texture analysis, due to its optimal localization properties in both spatial and frequency domain. Texture features based on Gabor wavelets transformation are found by the Gabor filtered image. Experiment results indicate the following: (1) the proposed algorithm based on random forest outperformed the-state-of-the-art in pigmented skin lesions detection (2) and the inclusion of Gabor wavelet transformation based texture features improved segmentation accuracy significantly.

Using classified Landsat Thematic Mapper data for stratification in a statewide forest inventory

Treesearch

Mark H. Hansen; Daniel G. Wendt

2000-01-01

The 1998 Indiana/Illinois forest inventory (USDA Forest Service, Forest Inventory and Analysis (FIA)) used Landsat Thematic Mapper (TM) data for stratification. Classified images made by the National Gap Analysis Program (GAP) stratified FIA plots into four classes (nonforest, nonforest/ forest, forest/nonforest, and forest) based on a two pixel forest edge buffer zone...

Using Classified Landsat Thematic Mapper Data for Stratification in a Statewide Forest Inventory

Treesearch

Mark H. Hansen; Daniel G. Wendt

2000-01-01

The 1998 Indiana/Illinois forest inventory (USDA Forest Service, Forest Inventory and Analysis (FIA)) used Landsat Thematic Mapper (TM} data for stratification. Classified images made by the National Gap Analysis Program (GAP) stratified FIA plots into four classes (nonforest, nonforest/forest, forest/nonforest, and forest) based on a two pixel forest edge buffer zone...

Analysis and Recognition of Traditional Chinese Medicine Pulse Based on the Hilbert-Huang Transform and Random Forest in Patients with Coronary Heart Disease

PubMed Central

Wang, Yiqin; Yan, Hanxia; Yan, Jianjun; Yuan, Fengyin; Xu, Zhaoxia; Liu, Guoping; Xu, Wenjie

2015-01-01

Objective. This research provides objective and quantitative parameters of the traditional Chinese medicine (TCM) pulse conditions for distinguishing between patients with the coronary heart disease (CHD) and normal people by using the proposed classification approach based on Hilbert-Huang transform (HHT) and random forest. Methods. The energy and the sample entropy features were extracted by applying the HHT to TCM pulse by treating these pulse signals as time series. By using the random forest classifier, the extracted two types of features and their combination were, respectively, used as input data to establish classification model. Results. Statistical results showed that there were significant differences in the pulse energy and sample entropy between the CHD group and the normal group. Moreover, the energy features, sample entropy features, and their combination were inputted as pulse feature vectors; the corresponding average recognition rates were 84%, 76.35%, and 90.21%, respectively. Conclusion. The proposed approach could be appropriately used to analyze pulses of patients with CHD, which can lay a foundation for research on objective and quantitative criteria on disease diagnosis or Zheng differentiation. PMID:26180536

Analysis and Recognition of Traditional Chinese Medicine Pulse Based on the Hilbert-Huang Transform and Random Forest in Patients with Coronary Heart Disease.

PubMed

Guo, Rui; Wang, Yiqin; Yan, Hanxia; Yan, Jianjun; Yuan, Fengyin; Xu, Zhaoxia; Liu, Guoping; Xu, Wenjie

2015-01-01

Objective. This research provides objective and quantitative parameters of the traditional Chinese medicine (TCM) pulse conditions for distinguishing between patients with the coronary heart disease (CHD) and normal people by using the proposed classification approach based on Hilbert-Huang transform (HHT) and random forest. Methods. The energy and the sample entropy features were extracted by applying the HHT to TCM pulse by treating these pulse signals as time series. By using the random forest classifier, the extracted two types of features and their combination were, respectively, used as input data to establish classification model. Results. Statistical results showed that there were significant differences in the pulse energy and sample entropy between the CHD group and the normal group. Moreover, the energy features, sample entropy features, and their combination were inputted as pulse feature vectors; the corresponding average recognition rates were 84%, 76.35%, and 90.21%, respectively. Conclusion. The proposed approach could be appropriately used to analyze pulses of patients with CHD, which can lay a foundation for research on objective and quantitative criteria on disease diagnosis or Zheng differentiation.

Automatic Estimation of Osteoporotic Fracture Cases by Using Ensemble Learning Approaches.

PubMed

Kilic, Niyazi; Hosgormez, Erkan

2016-03-01

Ensemble learning methods are one of the most powerful tools for the pattern classification problems. In this paper, the effects of ensemble learning methods and some physical bone densitometry parameters on osteoporotic fracture detection were investigated. Six feature set models were constructed including different physical parameters and they fed into the ensemble classifiers as input features. As ensemble learning techniques, bagging, gradient boosting and random subspace (RSM) were used. Instance based learning (IBk) and random forest (RF) classifiers applied to six feature set models. The patients were classified into three groups such as osteoporosis, osteopenia and control (healthy), using ensemble classifiers. Total classification accuracy and f-measure were also used to evaluate diagnostic performance of the proposed ensemble classification system. The classification accuracy has reached to 98.85 % by the combination of model 6 (five BMD + five T-score values) using RSM-RF classifier. The findings of this paper suggest that the patients will be able to be warned before a bone fracture occurred, by just examining some physical parameters that can easily be measured without invasive operations.

Large Unbalanced Credit Scoring Using Lasso-Logistic Regression Ensemble

PubMed Central

Wang, Hong; Xu, Qingsong; Zhou, Lifeng

2015-01-01

Recently, various ensemble learning methods with different base classifiers have been proposed for credit scoring problems. However, for various reasons, there has been little research using logistic regression as the base classifier. In this paper, given large unbalanced data, we consider the plausibility of ensemble learning using regularized logistic regression as the base classifier to deal with credit scoring problems. In this research, the data is first balanced and diversified by clustering and bagging algorithms. Then we apply a Lasso-logistic regression learning ensemble to evaluate the credit risks. We show that the proposed algorithm outperforms popular credit scoring models such as decision tree, Lasso-logistic regression and random forests in terms of AUC and F-measure. We also provide two importance measures for the proposed model to identify important variables in the data. PMID:25706988

Classification of California streams using combined deductive and inductive approaches: Setting the foundation for analysis of hydrologic alteration

USGS Publications Warehouse

Pyne, Matthew I.; Carlisle, Daren M.; Konrad, Christopher P.; Stein, Eric D.

2017-01-01

Regional classification of streams is an early step in the Ecological Limits of Hydrologic Alteration framework. Many stream classifications are based on an inductive approach using hydrologic data from minimally disturbed basins, but this approach may underrepresent streams from heavily disturbed basins or sparsely gaged arid regions. An alternative is a deductive approach, using watershed climate, land use, and geomorphology to classify streams, but this approach may miss important hydrological characteristics of streams. We classified all stream reaches in California using both approaches. First, we used Bayesian and hierarchical clustering to classify reaches according to watershed characteristics. Streams were clustered into seven classes according to elevation, sedimentary rock, and winter precipitation. Permutation-based analysis of variance and random forest analyses were used to determine which hydrologic variables best separate streams into their respective classes. Stream typology (i.e., the class that a stream reach is assigned to) is shaped mainly by patterns of high and mean flow behavior within the stream's landscape context. Additionally, random forest was used to determine which hydrologic variables best separate minimally disturbed reference streams from non-reference streams in each of the seven classes. In contrast to stream typology, deviation from reference conditions is more difficult to detect and is largely defined by changes in low-flow variables, average daily flow, and duration of flow. Our combined deductive/inductive approach allows us to estimate flow under minimally disturbed conditions based on the deductive analysis and compare to measured flow based on the inductive analysis in order to estimate hydrologic change.

Validation of psoriatic arthritis diagnoses in electronic medical records using natural language processing

PubMed Central

Cai, Tianxi; Karlson, Elizabeth W.

2013-01-01

Objectives To test whether data extracted from full text patient visit notes from an electronic medical record (EMR) would improve the classification of PsA compared to an algorithm based on codified data. Methods From the > 1,350,000 adults in a large academic EMR, all 2318 patients with a billing code for PsA were extracted and 550 were randomly selected for chart review and algorithm training. Using codified data and phrases extracted from narrative data using natural language processing, 31 predictors were extracted and three random forest algorithms trained using coded, narrative, and combined predictors. The receiver operator curve (ROC) was used to identify the optimal algorithm and a cut point was chosen to achieve the maximum sensitivity possible at a 90% positive predictive value (PPV). The algorithm was then used to classify the remaining 1768 charts and finally validated in a random sample of 300 cases predicted to have PsA. Results The PPV of a single PsA code was 57% (95%CI 55%–58%). Using a combination of coded data and NLP the random forest algorithm reached a PPV of 90% (95%CI 86%–93%) at sensitivity of 87% (95% CI 83% – 91%) in the training data. The PPV was 93% (95%CI 89%–96%) in the validation set. Adding NLP predictors to codified data increased the area under the ROC (p < 0.001). Conclusions Using NLP with text notes from electronic medical records improved the performance of the prediction algorithm significantly. Random forests were a useful tool to accurately classify psoriatic arthritis cases to enable epidemiological research. PMID:20701955

Source identification of western Oregon Douglas-fir wood cores using mass spectrometry and random forest classification1

PubMed Central

Finch, Kristen; Espinoza, Edgard; Jones, F. Andrew; Cronn, Richard

2017-01-01

Premise of the study: We investigated whether wood metabolite profiles from direct analysis in real time (time-of-flight) mass spectrometry (DART-TOFMS) could be used to determine the geographic origin of Douglas-fir wood cores originating from two regions in western Oregon, USA. Methods: Three annual ring mass spectra were obtained from 188 adult Douglas-fir trees, and these were analyzed using random forest models to determine whether samples could be classified to geographic origin, growth year, or growth year and geographic origin. Specific wood molecules that contributed to geographic discrimination were identified. Results: Douglas-fir mass spectra could be differentiated into two geographic classes with an accuracy between 70% and 76%. Classification models could not accurately classify sample mass spectra based on growth year. Thirty-two molecules were identified as key for classifying western Oregon Douglas-fir wood cores to geographic origin. Discussion: DART-TOFMS is capable of detecting minute but regionally informative differences in wood molecules over a small geographic scale, and these differences made it possible to predict the geographic origin of Douglas-fir wood with moderate accuracy. Studies involving DART-TOFMS, alone and in combination with other technologies, will be relevant for identifying the geographic origin of illegally harvested wood. PMID:28529831

Classification of Hyperspectral Data Based on Guided Filtering and Random Forest

NASA Astrophysics Data System (ADS)

Ma, H.; Feng, W.; Cao, X.; Wang, L.

2017-09-01

Hyperspectral images usually consist of more than one hundred spectral bands, which have potentials to provide rich spatial and spectral information. However, the application of hyperspectral data is still challengeable due to "the curse of dimensionality". In this context, many techniques, which aim to make full use of both the spatial and spectral information, are investigated. In order to preserve the geometrical information, meanwhile, with less spectral bands, we propose a novel method, which combines principal components analysis (PCA), guided image filtering and the random forest classifier (RF). In detail, PCA is firstly employed to reduce the dimension of spectral bands. Secondly, the guided image filtering technique is introduced to smooth land object, meanwhile preserving the edge of objects. Finally, the features are fed into RF classifier. To illustrate the effectiveness of the method, we carry out experiments over the popular Indian Pines data set, which is collected by Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) sensor. By comparing the proposed method with the method of only using PCA or guided image filter, we find that effect of the proposed method is better.

Automatic segmentation and classification of mycobacterium tuberculosis with conventional light microscopy

NASA Astrophysics Data System (ADS)

Xu, Chao; Zhou, Dongxiang; Zhai, Yongping; Liu, Yunhui

2015-12-01

This paper realizes the automatic segmentation and classification of Mycobacterium tuberculosis with conventional light microscopy. First, the candidate bacillus objects are segmented by the marker-based watershed transform. The markers are obtained by an adaptive threshold segmentation based on the adaptive scale Gaussian filter. The scale of the Gaussian filter is determined according to the color model of the bacillus objects. Then the candidate objects are extracted integrally after region merging and contaminations elimination. Second, the shape features of the bacillus objects are characterized by the Hu moments, compactness, eccentricity, and roughness, which are used to classify the single, touching and non-bacillus objects. We evaluated the logistic regression, random forest, and intersection kernel support vector machines classifiers in classifying the bacillus objects respectively. Experimental results demonstrate that the proposed method yields to high robustness and accuracy. The logistic regression classifier performs best with an accuracy of 91.68%.

Random forests in non-invasive sensorimotor rhythm brain-computer interfaces: a practical and convenient non-linear classifier.

PubMed

Steyrl, David; Scherer, Reinhold; Faller, Josef; Müller-Putz, Gernot R

2016-02-01

There is general agreement in the brain-computer interface (BCI) community that although non-linear classifiers can provide better results in some cases, linear classifiers are preferable. Particularly, as non-linear classifiers often involve a number of parameters that must be carefully chosen. However, new non-linear classifiers were developed over the last decade. One of them is the random forest (RF) classifier. Although popular in other fields of science, RFs are not common in BCI research. In this work, we address three open questions regarding RFs in sensorimotor rhythm (SMR) BCIs: parametrization, online applicability, and performance compared to regularized linear discriminant analysis (LDA). We found that the performance of RF is constant over a large range of parameter values. We demonstrate - for the first time - that RFs are applicable online in SMR-BCIs. Further, we show in an offline BCI simulation that RFs statistically significantly outperform regularized LDA by about 3%. These results confirm that RFs are practical and convenient non-linear classifiers for SMR-BCIs. Taking into account further properties of RFs, such as independence from feature distributions, maximum margin behavior, multiclass and advanced data mining capabilities, we argue that RFs should be taken into consideration for future BCIs.

Radiomic modeling of BI-RADS density categories

NASA Astrophysics Data System (ADS)

Wei, Jun; Chan, Heang-Ping; Helvie, Mark A.; Roubidoux, Marilyn A.; Zhou, Chuan; Hadjiiski, Lubomir

2017-03-01

Screening mammography is the most effective and low-cost method to date for early cancer detection. Mammographic breast density has been shown to be highly correlated with breast cancer risk. We are developing a radiomic model for BI-RADS density categorization on digital mammography (FFDM) with a supervised machine learning approach. With IRB approval, we retrospectively collected 478 FFDMs from 478 women. As a gold standard, breast density was assessed by an MQSA radiologist based on BI-RADS categories. The raw FFDMs were used for computerized density assessment. The raw FFDM first underwent log-transform to approximate the x-ray sensitometric response, followed by multiscale processing to enhance the fibroglandular densities and parenchymal patterns. Three ROIs were automatically identified based on the keypoint distribution, where the keypoints were obtained as the extrema in the image Gaussian scale-space. A total of 73 features, including intensity and texture features that describe the density and the parenchymal pattern, were extracted from each breast. Our BI-RADS density estimator was constructed by using a random forest classifier. We used a 10-fold cross validation resampling approach to estimate the errors. With the random forest classifier, computerized density categories for 412 of the 478 cases agree with radiologist's assessment (weighted kappa = 0.93). The machine learning method with radiomic features as predictors demonstrated a high accuracy in classifying FFDMs into BI-RADS density categories. Further work is underway to improve our system performance as well as to perform an independent testing using a large unseen FFDM set.

Comparative Performance Analysis of Support Vector Machine, Random Forest, Logistic Regression and k-Nearest Neighbours in Rainbow Trout (Oncorhynchus Mykiss) Classification Using Image-Based Features

PubMed Central

Císař, Petr; Labbé, Laurent; Souček, Pavel; Pelissier, Pablo; Kerneis, Thierry

2018-01-01

The main aim of this study was to develop a new objective method for evaluating the impacts of different diets on the live fish skin using image-based features. In total, one-hundred and sixty rainbow trout (Oncorhynchus mykiss) were fed either a fish-meal based diet (80 fish) or a 100% plant-based diet (80 fish) and photographed using consumer-grade digital camera. Twenty-three colour features and four texture features were extracted. Four different classification methods were used to evaluate fish diets including Random forest (RF), Support vector machine (SVM), Logistic regression (LR) and k-Nearest neighbours (k-NN). The SVM with radial based kernel provided the best classifier with correct classification rate (CCR) of 82% and Kappa coefficient of 0.65. Although the both LR and RF methods were less accurate than SVM, they achieved good classification with CCR 75% and 70% respectively. The k-NN was the least accurate (40%) classification model. Overall, it can be concluded that consumer-grade digital cameras could be employed as the fast, accurate and non-invasive sensor for classifying rainbow trout based on their diets. Furthermore, these was a close association between image-based features and fish diet received during cultivation. These procedures can be used as non-invasive, accurate and precise approaches for monitoring fish status during the cultivation by evaluating diet’s effects on fish skin. PMID:29596375

Comparative Performance Analysis of Support Vector Machine, Random Forest, Logistic Regression and k-Nearest Neighbours in Rainbow Trout (Oncorhynchus Mykiss) Classification Using Image-Based Features.

PubMed

Saberioon, Mohammadmehdi; Císař, Petr; Labbé, Laurent; Souček, Pavel; Pelissier, Pablo; Kerneis, Thierry

2018-03-29

The main aim of this study was to develop a new objective method for evaluating the impacts of different diets on the live fish skin using image-based features. In total, one-hundred and sixty rainbow trout ( Oncorhynchus mykiss ) were fed either a fish-meal based diet (80 fish) or a 100% plant-based diet (80 fish) and photographed using consumer-grade digital camera. Twenty-three colour features and four texture features were extracted. Four different classification methods were used to evaluate fish diets including Random forest (RF), Support vector machine (SVM), Logistic regression (LR) and k -Nearest neighbours ( k -NN). The SVM with radial based kernel provided the best classifier with correct classification rate (CCR) of 82% and Kappa coefficient of 0.65. Although the both LR and RF methods were less accurate than SVM, they achieved good classification with CCR 75% and 70% respectively. The k -NN was the least accurate (40%) classification model. Overall, it can be concluded that consumer-grade digital cameras could be employed as the fast, accurate and non-invasive sensor for classifying rainbow trout based on their diets. Furthermore, these was a close association between image-based features and fish diet received during cultivation. These procedures can be used as non-invasive, accurate and precise approaches for monitoring fish status during the cultivation by evaluating diet's effects on fish skin.

Evaluation of Classifier Performance for Multiclass Phenotype Discrimination in Untargeted Metabolomics.

PubMed

Trainor, Patrick J; DeFilippis, Andrew P; Rai, Shesh N

2017-06-21

Statistical classification is a critical component of utilizing metabolomics data for examining the molecular determinants of phenotypes. Despite this, a comprehensive and rigorous evaluation of the accuracy of classification techniques for phenotype discrimination given metabolomics data has not been conducted. We conducted such an evaluation using both simulated and real metabolomics datasets, comparing Partial Least Squares-Discriminant Analysis (PLS-DA), Sparse PLS-DA, Random Forests, Support Vector Machines (SVM), Artificial Neural Network, k -Nearest Neighbors ( k -NN), and Naïve Bayes classification techniques for discrimination. We evaluated the techniques on simulated data generated to mimic global untargeted metabolomics data by incorporating realistic block-wise correlation and partial correlation structures for mimicking the correlations and metabolite clustering generated by biological processes. Over the simulation studies, covariance structures, means, and effect sizes were stochastically varied to provide consistent estimates of classifier performance over a wide range of possible scenarios. The effects of the presence of non-normal error distributions, the introduction of biological and technical outliers, unbalanced phenotype allocation, missing values due to abundances below a limit of detection, and the effect of prior-significance filtering (dimension reduction) were evaluated via simulation. In each simulation, classifier parameters, such as the number of hidden nodes in a Neural Network, were optimized by cross-validation to minimize the probability of detecting spurious results due to poorly tuned classifiers. Classifier performance was then evaluated using real metabolomics datasets of varying sample medium, sample size, and experimental design. We report that in the most realistic simulation studies that incorporated non-normal error distributions, unbalanced phenotype allocation, outliers, missing values, and dimension reduction, classifier performance (least to greatest error) was ranked as follows: SVM, Random Forest, Naïve Bayes, sPLS-DA, Neural Networks, PLS-DA and k -NN classifiers. When non-normal error distributions were introduced, the performance of PLS-DA and k -NN classifiers deteriorated further relative to the remaining techniques. Over the real datasets, a trend of better performance of SVM and Random Forest classifier performance was observed.

Preliminary Evaluation of Methods for Classifying Forest Site Productivity Based on Species Composition in Western North Carolina

Treesearch

W. Henry McNab; F. Thomas Lloyd; David L. Loftis

2002-01-01

The species indicator approach to forest site classification was evaluated for 210 relatively undisturbed plots established by the USDA Forest Service Forest Inventory and Analysis uni (FIA) in western North Carolina. Plots were classified by low, medium, and high levels of productivity based on 10-year individual tree basal area increment data standardized for initial...

Can-Evo-Ens: Classifier stacking based evolutionary ensemble system for prediction of human breast cancer using amino acid sequences.

PubMed

Ali, Safdar; Majid, Abdul

2015-04-01

The diagnostic of human breast cancer is an intricate process and specific indicators may produce negative results. In order to avoid misleading results, accurate and reliable diagnostic system for breast cancer is indispensable. Recently, several interesting machine-learning (ML) approaches are proposed for prediction of breast cancer. To this end, we developed a novel classifier stacking based evolutionary ensemble system "Can-Evo-Ens" for predicting amino acid sequences associated with breast cancer. In this paper, first, we selected four diverse-type of ML algorithms of Naïve Bayes, K-Nearest Neighbor, Support Vector Machines, and Random Forest as base-level classifiers. These classifiers are trained individually in different feature spaces using physicochemical properties of amino acids. In order to exploit the decision spaces, the preliminary predictions of base-level classifiers are stacked. Genetic programming (GP) is then employed to develop a meta-classifier that optimal combine the predictions of the base classifiers. The most suitable threshold value of the best-evolved predictor is computed using Particle Swarm Optimization technique. Our experiments have demonstrated the robustness of Can-Evo-Ens system for independent validation dataset. The proposed system has achieved the highest value of Area Under Curve (AUC) of ROC Curve of 99.95% for cancer prediction. The comparative results revealed that proposed approach is better than individual ML approaches and conventional ensemble approaches of AdaBoostM1, Bagging, GentleBoost, and Random Subspace. It is expected that the proposed novel system would have a major impact on the fields of Biomedical, Genomics, Proteomics, Bioinformatics, and Drug Development. Copyright © 2015 Elsevier Inc. All rights reserved.

DeepDeath: Learning to predict the underlying cause of death with Big Data.

PubMed

Hassanzadeh, Hamid Reza; Ying Sha; Wang, May D

2017-07-01

Multiple cause-of-death data provides a valuable source of information that can be used to enhance health standards by predicting health related trajectories in societies with large populations. These data are often available in large quantities across U.S. states and require Big Data techniques to uncover complex hidden patterns. We design two different classes of models suitable for large-scale analysis of mortality data, a Hadoop-based ensemble of random forests trained over N-grams, and the DeepDeath, a deep classifier based on the recurrent neural network (RNN). We apply both classes to the mortality data provided by the National Center for Health Statistics and show that while both perform significantly better than the random classifier, the deep model that utilizes long short-term memory networks (LSTMs), surpasses the N-gram based models and is capable of learning the temporal aspect of the data without a need for building ad-hoc, expert-driven features.

Advances in SCA and RF-DNA Fingerprinting Through Enhanced Linear Regression Attacks and Application of Random Forest Classifiers

DTIC Science & Technology

2014-09-18

Converter AES Advance Encryption Standard ANN Artificial Neural Network APS Application Support AUC Area Under the Curve CPA Correlation Power Analysis ...Importance WGN White Gaussian Noise WPAN Wireless Personal Area Networks XEnv Cross-Environment XRx Cross-Receiver xxi ADVANCES IN SCA AND RF-DNA...based tool called KillerBee was released in 2009 that increases the exposure of ZigBee and other IEEE 802.15.4-based Wireless Personal Area Networks

Characterizing channel change along a multithread gravel-bed river using random forest image classification

NASA Astrophysics Data System (ADS)

Overstreet, B. T.; Legleiter, C. J.

2012-12-01

The Snake River in Grand Teton National Park is a dam-regulated but highly dynamic gravel-bed river that alternates between a single thread and a multithread planform. Identifying key drivers of channel change on this river could improve our understanding of 1) how flow regulation at Jackson Lake Dam has altered the character of the river over time; 2) how changes in the distribution of various types of vegetation impacts river dynamics; and 3) how the Snake River will respond to future human and climate driven disturbances. Despite the importance of monitoring planform changes over time, automated channel extraction and understanding the physical drivers contributing to channel change continue to be challenging yet critical steps in the remote sensing of riverine environments. In this study we use the random forest statistical technique to first classify land cover within the Snake River corridor and then extract channel features from a sequence of high-resolution multispectral images of the Snake River spanning the period from 2006 to 2012, which encompasses both exceptionally dry years and near-record runoff in 2011. We show that the random forest technique can be used to classify images with as few as four spectral bands with far greater accuracy than traditional single-tree classification approaches. Secondly, we couple random forest derived land cover maps with LiDAR derived topography, bathymetry, and canopy height to explore physical drivers contributing to observed channel changes on the Snake River. In conclusion we show that the random forest technique is a powerful tool for classifying multispectral images of rivers. Moreover, we hypothesize that with sufficient data for calculating spatially distributed metrics of channel form and more frequent channel monitoring, this tool can also be used to identify areas with high probabilities of channel change. Land cover maps of a portion of the Snake River produced from digital aerial photography from 2010 and a 2011 WorldView2 satellite image. This pair of maps thus captures changes that occurred during the 2011 runoff

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.

Fire detection system using random forest classification for image sequences of complex background

NASA Astrophysics Data System (ADS)

Kim, Onecue; Kang, Dong-Joong

2013-06-01

We present a fire alarm system based on image processing that detects fire accidents in various environments. To reduce false alarms that frequently appeared in earlier systems, we combined image features including color, motion, and blinking information. We specifically define the color conditions of fires in hue, saturation and value, and RGB color space. Fire features are represented as intensity variation, color mean and variance, motion, and image differences. Moreover, blinking fire features are modeled by using crossing patches. We propose an algorithm that classifies patches into fire or nonfire areas by using random forest supervised learning. We design an embedded surveillance device made with acrylonitrile butadiene styrene housing for stable fire detection in outdoor environments. The experimental results show that our algorithm works robustly in complex environments and is able to detect fires in real time.

Gray level co-occurrence and random forest algorithm-based gender determination with maxillary tooth plaster images.

PubMed

Akkoç, Betül; Arslan, Ahmet; Kök, Hatice

2016-06-01

Gender is one of the intrinsic properties of identity, with performance enhancement reducing the cluster when a search is performed. Teeth have durable and resistant structure, and as such are important sources of identification in disasters (accident, fire, etc.). In this study, gender determination is accomplished by maxillary tooth plaster models of 40 people (20 males and 20 females). The images of tooth plaster models are taken with a lighting mechanism set-up. A gray level co-occurrence matrix of the image with segmentation is formed and classified via a Random Forest (RF) algorithm by extracting pertinent features of the matrix. Automatic gender determination has a 90% success rate, with an applicable system to determine gender from maxillary tooth plaster images. Copyright © 2016 Elsevier Ltd. All rights reserved.

Classifying forest inventory data into species-based forest community types at broad extents: exploring tradeoffs among supervised and unsupervised approaches

Treesearch

Jennifer K. Costanza; Don Faber-Langendoen; John W. Coulston; David N. Wear

2018-01-01

Background: Knowledge of the different kinds of tree communities that currently exist can provide a baseline for assessing the ecological attributes of forests and monitoring future changes. Forest inventory data can facilitate the development of this baseline knowledge across broad extents, but they first must be classified into forest...

Toward Improving Electrocardiogram (ECG) Biometric Verification using Mobile Sensors: A Two-Stage Classifier Approach

PubMed Central

Tan, Robin; Perkowski, Marek

2017-01-01

Electrocardiogram (ECG) signals sensed from mobile devices pertain the potential for biometric identity recognition applicable in remote access control systems where enhanced data security is demanding. In this study, we propose a new algorithm that consists of a two-stage classifier combining random forest and wavelet distance measure through a probabilistic threshold schema, to improve the effectiveness and robustness of a biometric recognition system using ECG data acquired from a biosensor integrated into mobile devices. The proposed algorithm is evaluated using a mixed dataset from 184 subjects under different health conditions. The proposed two-stage classifier achieves a total of 99.52% subject verification accuracy, better than the 98.33% accuracy from random forest alone and 96.31% accuracy from wavelet distance measure algorithm alone. These results demonstrate the superiority of the proposed algorithm for biometric identification, hence supporting its practicality in areas such as cloud data security, cyber-security or remote healthcare systems. PMID:28230745

Toward Improving Electrocardiogram (ECG) Biometric Verification using Mobile Sensors: A Two-Stage Classifier Approach.

PubMed

Tan, Robin; Perkowski, Marek

2017-02-20

Electrocardiogram (ECG) signals sensed from mobile devices pertain the potential for biometric identity recognition applicable in remote access control systems where enhanced data security is demanding. In this study, we propose a new algorithm that consists of a two-stage classifier combining random forest and wavelet distance measure through a probabilistic threshold schema, to improve the effectiveness and robustness of a biometric recognition system using ECG data acquired from a biosensor integrated into mobile devices. The proposed algorithm is evaluated using a mixed dataset from 184 subjects under different health conditions. The proposed two-stage classifier achieves a total of 99.52% subject verification accuracy, better than the 98.33% accuracy from random forest alone and 96.31% accuracy from wavelet distance measure algorithm alone. These results demonstrate the superiority of the proposed algorithm for biometric identification, hence supporting its practicality in areas such as cloud data security, cyber-security or remote healthcare systems.

Computer aided diagnosis system for the Alzheimer's disease based on partial least squares and random forest SPECT image classification.

PubMed

Ramírez, J; Górriz, J M; Segovia, F; Chaves, R; Salas-Gonzalez, D; López, M; Alvarez, I; Padilla, P

2010-03-19

This letter shows a computer aided diagnosis (CAD) technique for the early detection of the Alzheimer's disease (AD) by means of single photon emission computed tomography (SPECT) image classification. The proposed method is based on partial least squares (PLS) regression model and a random forest (RF) predictor. The challenge of the curse of dimensionality is addressed by reducing the large dimensionality of the input data by downscaling the SPECT images and extracting score features using PLS. A RF predictor then forms an ensemble of classification and regression tree (CART)-like classifiers being its output determined by a majority vote of the trees in the forest. A baseline principal component analysis (PCA) system is also developed for reference. The experimental results show that the combined PLS-RF system yields a generalization error that converges to a limit when increasing the number of trees in the forest. Thus, the generalization error is reduced when using PLS and depends on the strength of the individual trees in the forest and the correlation between them. Moreover, PLS feature extraction is found to be more effective for extracting discriminative information from the data than PCA yielding peak sensitivity, specificity and accuracy values of 100%, 92.7%, and 96.9%, respectively. Moreover, the proposed CAD system outperformed several other recently developed AD CAD systems. Copyright 2010 Elsevier Ireland Ltd. All rights reserved.

Random Forests Are Able to Identify Differences in Clotting Dynamics from Kinetic Models of Thrombin Generation.

PubMed

Arumugam, Jayavel; Bukkapatnam, Satish T S; Narayanan, Krishna R; Srinivasa, Arun R

2016-01-01

Current methods for distinguishing acute coronary syndromes such as heart attack from stable coronary artery disease, based on the kinetics of thrombin formation, have been limited to evaluating sensitivity of well-established chemical species (e.g., thrombin) using simple quantifiers of their concentration profiles (e.g., maximum level of thrombin concentration, area under the thrombin concentration versus time curve). In order to get an improved classifier, we use a 34-protein factor clotting cascade model and convert the simulation data into a high-dimensional representation (about 19000 features) using a piecewise cubic polynomial fit. Then, we systematically find plausible assays to effectively gauge changes in acute coronary syndrome/coronary artery disease populations by introducing a statistical learning technique called Random Forests. We find that differences associated with acute coronary syndromes emerge in combinations of a handful of features. For instance, concentrations of 3 chemical species, namely, active alpha-thrombin, tissue factor-factor VIIa-factor Xa ternary complex, and intrinsic tenase complex with factor X, at specific time windows, could be used to classify acute coronary syndromes to an accuracy of about 87.2%. Such a combination could be used to efficiently assay the coagulation system.

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.

[Object-oriented segmentation and classification of forest gap based on QuickBird remote sensing image.

PubMed

Mao, Xue Gang; Du, Zi Han; Liu, Jia Qian; Chen, Shu Xin; Hou, Ji Yu

2018-01-01

Traditional field investigation and artificial interpretation could not satisfy the need of forest gaps extraction at regional scale. High spatial resolution remote sensing image provides the possibility for regional forest gaps extraction. In this study, we used object-oriented classification method to segment and classify forest gaps based on QuickBird high resolution optical remote sensing image in Jiangle National Forestry Farm of Fujian Province. In the process of object-oriented classification, 10 scales (10-100, with a step length of 10) were adopted to segment QuickBird remote sensing image; and the intersection area of reference object (RA or ) and intersection area of segmented object (RA os ) were adopted to evaluate the segmentation result at each scale. For segmentation result at each scale, 16 spectral characteristics and support vector machine classifier (SVM) were further used to classify forest gaps, non-forest gaps and others. The results showed that the optimal segmentation scale was 40 when RA or was equal to RA os . The accuracy difference between the maximum and minimum at different segmentation scales was 22%. At optimal scale, the overall classification accuracy was 88% (Kappa=0.82) based on SVM classifier. Combining high resolution remote sensing image data with object-oriented classification method could replace the traditional field investigation and artificial interpretation method to identify and classify forest gaps at regional scale.

Using random forest for reliable classification and cost-sensitive learning for medical diagnosis.

PubMed

Yang, Fan; Wang, Hua-zhen; Mi, Hong; Lin, Cheng-de; Cai, Wei-wen

2009-01-30

Most machine-learning classifiers output label predictions for new instances without indicating how reliable the predictions are. The applicability of these classifiers is limited in critical domains where incorrect predictions have serious consequences, like medical diagnosis. Further, the default assumption of equal misclassification costs is most likely violated in medical diagnosis. In this paper, we present a modified random forest classifier which is incorporated into the conformal predictor scheme. A conformal predictor is a transductive learning scheme, using Kolmogorov complexity to test the randomness of a particular sample with respect to the training sets. Our method show well-calibrated property that the performance can be set prior to classification and the accurate rate is exactly equal to the predefined confidence level. Further, to address the cost sensitive problem, we extend our method to a label-conditional predictor which takes into account different costs for misclassifications in different class and allows different confidence level to be specified for each class. Intensive experiments on benchmark datasets and real world applications show the resultant classifier is well-calibrated and able to control the specific risk of different class. The method of using RF outlier measure to design a nonconformity measure benefits the resultant predictor. Further, a label-conditional classifier is developed and turn to be an alternative approach to the cost sensitive learning problem that relies on label-wise predefined confidence level. The target of minimizing the risk of misclassification is achieved by specifying the different confidence level for different class.

Identification of terrain cover using the optimum polarimetric classifier

NASA Technical Reports Server (NTRS)

Kong, J. A.; Swartz, A. A.; Yueh, H. A.; Novak, L. M.; Shin, R. T.

1988-01-01

A systematic approach for the identification of terrain media such as vegetation canopy, forest, and snow-covered fields is developed using the optimum polarimetric classifier. The covariance matrices for various terrain cover are computed from theoretical models of random medium by evaluating the scattering matrix elements. The optimal classification scheme makes use of a quadratic distance measure and is applied to classify a vegetation canopy consisting of both trees and grass. Experimentally measured data are used to validate the classification scheme. Analytical and Monte Carlo simulated classification errors using the fully polarimetric feature vector are compared with classification based on single features which include the phase difference between the VV and HH polarization returns. It is shown that the full polarimetric results are optimal and provide better classification performance than single feature measurements.

Evolving optimised decision rules for intrusion detection using particle swarm paradigm

NASA Astrophysics Data System (ADS)

Sivatha Sindhu, Siva S.; Geetha, S.; Kannan, A.

2012-12-01

The aim of this article is to construct a practical intrusion detection system (IDS) that properly analyses the statistics of network traffic pattern and classify them as normal or anomalous class. The objective of this article is to prove that the choice of effective network traffic features and a proficient machine-learning paradigm enhances the detection accuracy of IDS. In this article, a rule-based approach with a family of six decision tree classifiers, namely Decision Stump, C4.5, Naive Baye's Tree, Random Forest, Random Tree and Representative Tree model to perform the detection of anomalous network pattern is introduced. In particular, the proposed swarm optimisation-based approach selects instances that compose training set and optimised decision tree operate over this trained set producing classification rules with improved coverage, classification capability and generalisation ability. Experiment with the Knowledge Discovery and Data mining (KDD) data set which have information on traffic pattern, during normal and intrusive behaviour shows that the proposed algorithm produces optimised decision rules and outperforms other machine-learning algorithm.

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.

A Hybrid Color Space for Skin Detection Using Genetic Algorithm Heuristic Search and Principal Component Analysis Technique

PubMed Central

2015-01-01

Color is one of the most prominent features of an image and used in many skin and face detection applications. Color space transformation is widely used by researchers to improve face and skin detection performance. Despite the substantial research efforts in this area, choosing a proper color space in terms of skin and face classification performance which can address issues like illumination variations, various camera characteristics and diversity in skin color tones has remained an open issue. This research proposes a new three-dimensional hybrid color space termed SKN by employing the Genetic Algorithm heuristic and Principal Component Analysis to find the optimal representation of human skin color in over seventeen existing color spaces. Genetic Algorithm heuristic is used to find the optimal color component combination setup in terms of skin detection accuracy while the Principal Component Analysis projects the optimal Genetic Algorithm solution to a less complex dimension. Pixel wise skin detection was used to evaluate the performance of the proposed color space. We have employed four classifiers including Random Forest, Naïve Bayes, Support Vector Machine and Multilayer Perceptron in order to generate the human skin color predictive model. The proposed color space was compared to some existing color spaces and shows superior results in terms of pixel-wise skin detection accuracy. Experimental results show that by using Random Forest classifier, the proposed SKN color space obtained an average F-score and True Positive Rate of 0.953 and False Positive Rate of 0.0482 which outperformed the existing color spaces in terms of pixel wise skin detection accuracy. The results also indicate that among the classifiers used in this study, Random Forest is the most suitable classifier for pixel wise skin detection applications. PMID:26267377

Recognizing pedestrian's unsafe behaviors in far-infrared imagery at night

NASA Astrophysics Data System (ADS)

Lee, Eun Ju; Ko, Byoung Chul; Nam, Jae-Yeal

2016-05-01

Pedestrian behavior recognition is important work for early accident prevention in advanced driver assistance system (ADAS). In particular, because most pedestrian-vehicle crashes are occurred from late of night to early of dawn, our study focus on recognizing unsafe behavior of pedestrians using thermal image captured from moving vehicle at night. For recognizing unsafe behavior, this study uses convolutional neural network (CNN) which shows high quality of recognition performance. However, because traditional CNN requires the very expensive training time and memory, we design the light CNN consisted of two convolutional layers and two subsampling layers for real-time processing of vehicle applications. In addition, we combine light CNN with boosted random forest (Boosted RF) classifier so that the output of CNN is not fully connected with the classifier but randomly connected with Boosted random forest. We named this CNN as randomly connected CNN (RC-CNN). The proposed method was successfully applied to the pedestrian unsafe behavior (PUB) dataset captured from far-infrared camera at night and its behavior recognition accuracy is confirmed to be higher than that of some algorithms related to CNNs, with a shorter processing time.

Empirical Evaluation of Hunk Metrics as Bug Predictors

NASA Astrophysics Data System (ADS)

Ferzund, Javed; Ahsan, Syed Nadeem; Wotawa, Franz

Reducing the number of bugs is a crucial issue during software development and maintenance. Software process and product metrics are good indicators of software complexity. These metrics have been used to build bug predictor models to help developers maintain the quality of software. In this paper we empirically evaluate the use of hunk metrics as predictor of bugs. We present a technique for bug prediction that works at smallest units of code change called hunks. We build bug prediction models using random forests, which is an efficient machine learning classifier. Hunk metrics are used to train the classifier and each hunk metric is evaluated for its bug prediction capabilities. Our classifier can classify individual hunks as buggy or bug-free with 86 % accuracy, 83 % buggy hunk precision and 77% buggy hunk recall. We find that history based and change level hunk metrics are better predictors of bugs than code level hunk metrics.

MOWGLI: prediction of protein-MannOse interacting residues With ensemble classifiers usinG evoLutionary Information.

PubMed

Pai, Priyadarshini P; Mondal, Sukanta

2016-10-01

Proteins interact with carbohydrates to perform various cellular interactions. Of the many carbohydrate ligands that proteins bind with, mannose constitute an important class, playing important roles in host defense mechanisms. Accurate identification of mannose-interacting residues (MIR) may provide important clues to decipher the underlying mechanisms of protein-mannose interactions during infections. This study proposes an approach using an ensemble of base classifiers for prediction of MIR using their evolutionary information in the form of position-specific scoring matrix. The base classifiers are random forests trained by different subsets of training data set Dset128 using 10-fold cross-validation. The optimized ensemble of base classifiers, MOWGLI, is then used to predict MIR on protein chains of the test data set Dtestset29 which showed a promising performance with 92.0% accurate prediction. An overall improvement of 26.6% in precision was observed upon comparison with the state-of-art. It is hoped that this approach, yielding enhanced predictions, could be eventually used for applications in drug design and vaccine development.

Predicting healthcare associated infections using patients' experiences

NASA Astrophysics Data System (ADS)

Pratt, Michael A.; Chu, Henry

2016-05-01

Healthcare associated infections (HAI) are a major threat to patient safety and are costly to health systems. Our goal is to predict the HAI performance of a hospital using the patients' experience responses as input. We use four classifiers, viz. random forest, naive Bayes, artificial feedforward neural networks, and the support vector machine, to perform the prediction of six types of HAI. The six types include blood stream, urinary tract, surgical site, and intestinal infections. Experiments show that the random forest and support vector machine perform well across the six types of HAI.

Preliminary classification of forest vegetation of the Kenai Peninsula, Alaska.

Treesearch

K.M. Reynolds

1990-01-01

A total of 5,597 photo points was systematically located on 1:60,000-scale high altitude photographs of the Kenai Peninsula, Alaska; photo interpretation was used to classify the vegetation at each grid position. Of the total grid points, 12.3 percent were classified as timberland; 129 photo points within the timberland class were randomly selected for field survey....

Filling of Cloud-Induced Gaps for Land Use and Land Cover Classifications Around Refugee Camps

NASA Astrophysics Data System (ADS)

Braun, Andreas; Hagensieker, Ron; Hochschild, Volker

2016-08-01

Clouds cover is one of the main constraints in the field of optical remote sensing. Especially the use of multispectral imagery is affected by either fully obscured data or parts of the image which remain unusable. This study compares four algorithms for the filling of cloud induced gaps in classified land cover products based on Markov Random Fields (MRF), Random Forest (RF), Closest Spectral Fit (CSF) operators. They are tested on a classified image of Sentinel-2 where artificial clouds are filled by information derived from a scene of Sentinel-1. The approaches rely on different mathematical principles and therefore produced results varying in both pattern and quality. Overall accuracies for the filled areas range from 57 to 64 %. Best results are achieved by CSF, however some classes (e.g. sands and grassland) remain critical through all approaches.

Random forest classification of large volume structures for visuo-haptic rendering in CT images

NASA Astrophysics Data System (ADS)

Mastmeyer, Andre; Fortmeier, Dirk; Handels, Heinz

2016-03-01

For patient-specific voxel-based visuo-haptic rendering of CT scans of the liver area, the fully automatic segmentation of large volume structures such as skin, soft tissue, lungs and intestine (risk structures) is important. Using a machine learning based approach, several existing segmentations from 10 segmented gold-standard patients are learned by random decision forests individually and collectively. The core of this paper is feature selection and the application of the learned classifiers to a new patient data set. In a leave-some-out cross-validation, the obtained full volume segmentations are compared to the gold-standard segmentations of the untrained patients. The proposed classifiers use a multi-dimensional feature space to estimate the hidden truth, instead of relying on clinical standard threshold and connectivity based methods. The result of our efficient whole-body section classification are multi-label maps with the considered tissues. For visuo-haptic simulation, other small volume structures would have to be segmented additionally. We also take a look into these structures (liver vessels). For an experimental leave-some-out study consisting of 10 patients, the proposed method performs much more efficiently compared to state of the art methods. In two variants of leave-some-out experiments we obtain best mean DICE ratios of 0.79, 0.97, 0.63 and 0.83 for skin, soft tissue, hard bone and risk structures. Liver structures are segmented with DICE 0.93 for the liver, 0.43 for blood vessels and 0.39 for bile vessels.

Testing random forest classification for identifying lava flows and mapping age groups on a single Landsat 8 image

NASA Astrophysics Data System (ADS)

Li, Long; Solana, Carmen; Canters, Frank; Kervyn, Matthieu

2017-10-01

Mapping lava flows using satellite images is an important application of remote sensing in volcanology. Several volcanoes have been mapped through remote sensing using a wide range of data, from optical to thermal infrared and radar images, using techniques such as manual mapping, supervised/unsupervised classification, and elevation subtraction. So far, spectral-based mapping applications mainly focus on the use of traditional pixel-based classifiers, without much investigation into the added value of object-based approaches and into advantages of using machine learning algorithms. In this study, Nyamuragira, characterized by a series of > 20 overlapping lava flows erupted over the last century, was used as a case study. The random forest classifier was tested to map lava flows based on pixels and objects. Image classification was conducted for the 20 individual flows and for 8 groups of flows of similar age using a Landsat 8 image and a DEM of the volcano, both at 30-meter spatial resolution. Results show that object-based classification produces maps with continuous and homogeneous lava surfaces, in agreement with the physical characteristics of lava flows, while lava flows mapped through the pixel-based classification are heterogeneous and fragmented including much "salt and pepper noise". In terms of accuracy, both pixel-based and object-based classification performs well but the former results in higher accuracies than the latter except for mapping lava flow age groups without using topographic features. It is concluded that despite spectral similarity, lava flows of contrasting age can be well discriminated and mapped by means of image classification. The classification approach demonstrated in this study only requires easily accessible image data and can be applied to other volcanoes as well if there is sufficient information to calibrate the mapping.

Building Diversified Multiple Trees for classification in high dimensional noisy biomedical data.

PubMed

Li, Jiuyong; Liu, Lin; Liu, Jixue; Green, Ryan

2017-12-01

It is common that a trained classification model is applied to the operating data that is deviated from the training data because of noise. This paper will test an ensemble method, Diversified Multiple Tree (DMT), on its capability for classifying instances in a new laboratory using the classifier built on the instances of another laboratory. DMT is tested on three real world biomedical data sets from different laboratories in comparison with four benchmark ensemble methods, AdaBoost, Bagging, Random Forests, and Random Trees. Experiments have also been conducted on studying the limitation of DMT and its possible variations. Experimental results show that DMT is significantly more accurate than other benchmark ensemble classifiers on classifying new instances of a different laboratory from the laboratory where instances are used to build the classifier. This paper demonstrates that an ensemble classifier, DMT, is more robust in classifying noisy data than other widely used ensemble methods. DMT works on the data set that supports multiple simple trees.

Using random forests for assistance in the curation of G-protein coupled receptor databases.

PubMed

Shkurin, Aleksei; Vellido, Alfredo

2017-08-18

Biology is experiencing a gradual but fast transformation from a laboratory-centred science towards a data-centred one. As such, it requires robust data engineering and the use of quantitative data analysis methods as part of database curation. This paper focuses on G protein-coupled receptors, a large and heterogeneous super-family of cell membrane proteins of interest to biology in general. One of its families, Class C, is of particular interest to pharmacology and drug design. This family is quite heterogeneous on its own, and the discrimination of its several sub-families is a challenging problem. In the absence of known crystal structure, such discrimination must rely on their primary amino acid sequences. We are interested not as much in achieving maximum sub-family discrimination accuracy using quantitative methods, but in exploring sequence misclassification behavior. Specifically, we are interested in isolating those sequences showing consistent misclassification, that is, sequences that are very often misclassified and almost always to the same wrong sub-family. Random forests are used for this analysis due to their ensemble nature, which makes them naturally suited to gauge the consistency of misclassification. This consistency is here defined through the voting scheme of their base tree classifiers. Detailed consistency results for the random forest ensemble classification were obtained for all receptors and for all data transformations of their unaligned primary sequences. Shortlists of the most consistently misclassified receptors for each subfamily and transformation, as well as an overall shortlist including those cases that were consistently misclassified across transformations, were obtained. The latter should be referred to experts for further investigation as a data curation task. The automatic discrimination of the Class C sub-families of G protein-coupled receptors from their unaligned primary sequences shows clear limits. This study has investigated in some detail the consistency of their misclassification using random forest ensemble classifiers. Different sub-families have been shown to display very different discrimination consistency behaviors. The individual identification of consistently misclassified sequences should provide a tool for quality control to GPCR database curators.

Fast image interpolation via random forests.

PubMed

Huang, Jun-Jie; Siu, Wan-Chi; Liu, Tian-Rui

2015-10-01

This paper proposes a two-stage framework for fast image interpolation via random forests (FIRF). The proposed FIRF method gives high accuracy, as well as requires low computation. The underlying idea of this proposed work is to apply random forests to classify the natural image patch space into numerous subspaces and learn a linear regression model for each subspace to map the low-resolution image patch to high-resolution image patch. The FIRF framework consists of two stages. Stage 1 of the framework removes most of the ringing and aliasing artifacts in the initial bicubic interpolated image, while Stage 2 further refines the Stage 1 interpolated image. By varying the number of decision trees in the random forests and the number of stages applied, the proposed FIRF method can realize computationally scalable image interpolation. Extensive experimental results show that the proposed FIRF(3, 2) method achieves more than 0.3 dB improvement in peak signal-to-noise ratio over the state-of-the-art nonlocal autoregressive modeling (NARM) method. Moreover, the proposed FIRF(1, 1) obtains similar or better results as NARM while only takes its 0.3% computational time.

Application of random forests methods to diabetic retinopathy classification analyses.

PubMed

Casanova, Ramon; Saldana, Santiago; Chew, Emily Y; Danis, Ronald P; Greven, Craig M; Ambrosius, Walter T

2014-01-01

Diabetic retinopathy (DR) is one of the leading causes of blindness in the United States and world-wide. DR is a silent disease that may go unnoticed until it is too late for effective treatment. Therefore, early detection could improve the chances of therapeutic interventions that would alleviate its effects. Graded fundus photography and systemic data from 3443 ACCORD-Eye Study participants were used to estimate Random Forest (RF) and logistic regression classifiers. We studied the impact of sample size on classifier performance and the possibility of using RF generated class conditional probabilities as metrics describing DR risk. RF measures of variable importance are used to detect factors that affect classification performance. Both types of data were informative when discriminating participants with or without DR. RF based models produced much higher classification accuracy than those based on logistic regression. Combining both types of data did not increase accuracy but did increase statistical discrimination of healthy participants who subsequently did or did not have DR events during four years of follow-up. RF variable importance criteria revealed that microaneurysms counts in both eyes seemed to play the most important role in discrimination among the graded fundus variables, while the number of medicines and diabetes duration were the most relevant among the systemic variables. We have introduced RF methods to DR classification analyses based on fundus photography data. In addition, we propose an approach to DR risk assessment based on metrics derived from graded fundus photography and systemic data. Our results suggest that RF methods could be a valuable tool to diagnose DR diagnosis and evaluate its progression.

Human tracking in thermal images using adaptive particle filters with online random forest learning

NASA Astrophysics Data System (ADS)

Ko, Byoung Chul; Kwak, Joon-Young; Nam, Jae-Yeal

2013-11-01

This paper presents a fast and robust human tracking method to use in a moving long-wave infrared thermal camera under poor illumination with the existence of shadows and cluttered backgrounds. To improve the human tracking performance while minimizing the computation time, this study proposes an online learning of classifiers based on particle filters and combination of a local intensity distribution (LID) with oriented center-symmetric local binary patterns (OCS-LBP). Specifically, we design a real-time random forest (RF), which is the ensemble of decision trees for confidence estimation, and confidences of the RF are converted into a likelihood function of the target state. First, the target model is selected by the user and particles are sampled. Then, RFs are generated using the positive and negative examples with LID and OCS-LBP features by online learning. The learned RF classifiers are used to detect the most likely target position in the subsequent frame in the next stage. Then, the RFs are learned again by means of fast retraining with the tracked object and background appearance in the new frame. The proposed algorithm is successfully applied to various thermal videos as tests and its tracking performance is better than those of other methods.

Application of Machine Learning Approaches for Classifying Sitting Posture Based on Force and Acceleration Sensors.

PubMed

Zemp, Roland; Tanadini, Matteo; Plüss, Stefan; Schnüriger, Karin; Singh, Navrag B; Taylor, William R; Lorenzetti, Silvio

2016-01-01

Occupational musculoskeletal disorders, particularly chronic low back pain (LBP), are ubiquitous due to prolonged static sitting or nonergonomic sitting positions. Therefore, the aim of this study was to develop an instrumented chair with force and acceleration sensors to determine the accuracy of automatically identifying the user's sitting position by applying five different machine learning methods (Support Vector Machines, Multinomial Regression, Boosting, Neural Networks, and Random Forest). Forty-one subjects were requested to sit four times in seven different prescribed sitting positions (total 1148 samples). Sixteen force sensor values and the backrest angle were used as the explanatory variables (features) for the classification. The different classification methods were compared by means of a Leave-One-Out cross-validation approach. The best performance was achieved using the Random Forest classification algorithm, producing a mean classification accuracy of 90.9% for subjects with which the algorithm was not familiar. The classification accuracy varied between 81% and 98% for the seven different sitting positions. The present study showed the possibility of accurately classifying different sitting positions by means of the introduced instrumented office chair combined with machine learning analyses. The use of such novel approaches for the accurate assessment of chair usage could offer insights into the relationships between sitting position, sitting behaviour, and the occurrence of musculoskeletal disorders.

How random is the random forest? Random forest algorithm on the service of structural imaging biomarkers for Alzheimer's disease: from Alzheimer's disease neuroimaging initiative (ADNI) database.

PubMed

Dimitriadis, Stavros I; Liparas, Dimitris

2018-06-01

Neuroinformatics is a fascinating research field that applies computational models and analytical tools to high dimensional experimental neuroscience data for a better understanding of how the brain functions or dysfunctions in brain diseases. Neuroinformaticians work in the intersection of neuroscience and informatics supporting the integration of various sub-disciplines (behavioural neuroscience, genetics, cognitive psychology, etc.) working on brain research. Neuroinformaticians are the pathway of information exchange between informaticians and clinicians for a better understanding of the outcome of computational models and the clinical interpretation of the analysis. Machine learning is one of the most significant computational developments in the last decade giving tools to neuroinformaticians and finally to radiologists and clinicians for an automatic and early diagnosis-prognosis of a brain disease. Random forest (RF) algorithm has been successfully applied to high-dimensional neuroimaging data for feature reduction and also has been applied to classify the clinical label of a subject using single or multi-modal neuroimaging datasets. Our aim was to review the studies where RF was applied to correctly predict the Alzheimer's disease (AD), the conversion from mild cognitive impairment (MCI) and its robustness to overfitting, outliers and handling of non-linear data. Finally, we described our RF-based model that gave us the 1 st position in an international challenge for automated prediction of MCI from MRI data.

Ensemble Pruning for Glaucoma Detection in an Unbalanced Data Set.

PubMed

Adler, Werner; Gefeller, Olaf; Gul, Asma; Horn, Folkert K; Khan, Zardad; Lausen, Berthold

2016-12-07

Random forests are successful classifier ensemble methods consisting of typically 100 to 1000 classification trees. Ensemble pruning techniques reduce the computational cost, especially the memory demand, of random forests by reducing the number of trees without relevant loss of performance or even with increased performance of the sub-ensemble. The application to the problem of an early detection of glaucoma, a severe eye disease with low prevalence, based on topographical measurements of the eye background faces specific challenges. We examine the performance of ensemble pruning strategies for glaucoma detection in an unbalanced data situation. The data set consists of 102 topographical features of the eye background of 254 healthy controls and 55 glaucoma patients. We compare the area under the receiver operating characteristic curve (AUC), and the Brier score on the total data set, in the majority class, and in the minority class of pruned random forest ensembles obtained with strategies based on the prediction accuracy of greedily grown sub-ensembles, the uncertainty weighted accuracy, and the similarity between single trees. To validate the findings and to examine the influence of the prevalence of glaucoma in the data set, we additionally perform a simulation study with lower prevalences of glaucoma. In glaucoma classification all three pruning strategies lead to improved AUC and smaller Brier scores on the total data set with sub-ensembles as small as 30 to 80 trees compared to the classification results obtained with the full ensemble consisting of 1000 trees. In the simulation study, we were able to show that the prevalence of glaucoma is a critical factor and lower prevalence decreases the performance of our pruning strategies. The memory demand for glaucoma classification in an unbalanced data situation based on random forests could effectively be reduced by the application of pruning strategies without loss of performance in a population with increased risk of glaucoma.

Multiple-Primitives Hierarchical Classification of Airborne Laser Scanning Data in Urban Areas

NASA Astrophysics Data System (ADS)

Ni, H.; Lin, X. G.; Zhang, J. X.

2017-09-01

A hierarchical classification method for Airborne Laser Scanning (ALS) data of urban areas is proposed in this paper. This method is composed of three stages among which three types of primitives are utilized, i.e., smooth surface, rough surface, and individual point. In the first stage, the input ALS data is divided into smooth surfaces and rough surfaces by employing a step-wise point cloud segmentation method. In the second stage, classification based on smooth surfaces and rough surfaces is performed. Points in the smooth surfaces are first classified into ground and buildings based on semantic rules. Next, features of rough surfaces are extracted. Then, points in rough surfaces are classified into vegetation and vehicles based on the derived features and Random Forests (RF). In the third stage, point-based features are extracted for the ground points, and then, an individual point classification procedure is performed to classify the ground points into bare land, artificial ground and greenbelt. Moreover, the shortages of the existing studies are analyzed, and experiments show that the proposed method overcomes these shortages and handles more types of objects.

Sequential Monte Carlo tracking of the marginal artery by multiple cue fusion and random forest regression.

PubMed

Cherry, Kevin M; Peplinski, Brandon; Kim, Lauren; Wang, Shijun; Lu, Le; Zhang, Weidong; Liu, Jianfei; Wei, Zhuoshi; Summers, Ronald M

2015-01-01

Given the potential importance of marginal artery localization in automated registration in computed tomography colonography (CTC), we have devised a semi-automated method of marginal vessel detection employing sequential Monte Carlo tracking (also known as particle filtering tracking) by multiple cue fusion based on intensity, vesselness, organ detection, and minimum spanning tree information for poorly enhanced vessel segments. We then employed a random forest algorithm for intelligent cue fusion and decision making which achieved high sensitivity and robustness. After applying a vessel pruning procedure to the tracking results, we achieved statistically significantly improved precision compared to a baseline Hessian detection method (2.7% versus 75.2%, p<0.001). This method also showed statistically significantly improved recall rate compared to a 2-cue baseline method using fewer vessel cues (30.7% versus 67.7%, p<0.001). These results demonstrate that marginal artery localization on CTC is feasible by combining a discriminative classifier (i.e., random forest) with a sequential Monte Carlo tracking mechanism. In so doing, we present the effective application of an anatomical probability map to vessel pruning as well as a supplementary spatial coordinate system for colonic segmentation and registration when this task has been confounded by colon lumen collapse. Published by Elsevier B.V.

Integrating Geo-Spatial Data for Regional Landslide Susceptibility Modeling in Consideration of Run-Out Signature

NASA Astrophysics Data System (ADS)

Lai, J.-S.; Tsai, F.; Chiang, S.-H.

2016-06-01

This study implements a data mining-based algorithm, the random forests classifier, with geo-spatial data to construct a regional and rainfall-induced landslide susceptibility model. The developed model also takes account of landslide regions (source, non-occurrence and run-out signatures) from the original landslide inventory in order to increase the reliability of the susceptibility modelling. A total of ten causative factors were collected and used in this study, including aspect, curvature, elevation, slope, faults, geology, NDVI (Normalized Difference Vegetation Index), rivers, roads and soil data. Consequently, this study transforms the landslide inventory and vector-based causative factors into the pixel-based format in order to overlay with other raster data for constructing the random forests based model. This study also uses original and edited topographic data in the analysis to understand their impacts to the susceptibility modeling. Experimental results demonstrate that after identifying the run-out signatures, the overall accuracy and Kappa coefficient have been reached to be become more than 85 % and 0.8, respectively. In addition, correcting unreasonable topographic feature of the digital terrain model also produces more reliable modelling results.

Fangorn Forest (F2): a machine learning approach to classify genes and genera in the family Geminiviridae.

PubMed

Silva, José Cleydson F; Carvalho, Thales F M; Fontes, Elizabeth P B; Cerqueira, Fabio R

2017-09-30

Geminiviruses infect a broad range of cultivated and non-cultivated plants, causing significant economic losses worldwide. The studies of the diversity of species, taxonomy, mechanisms of evolution, geographic distribution, and mechanisms of interaction of these pathogens with the host have greatly increased in recent years. Furthermore, the use of rolling circle amplification (RCA) and advanced metagenomics approaches have enabled the elucidation of viromes and the identification of many viral agents in a large number of plant species. As a result, determining the nomenclature and taxonomically classifying geminiviruses turned into complex tasks. In addition, the gene responsible for viral replication (particularly, the viruses belonging to the genus Mastrevirus) may be spliced due to the use of the transcriptional/splicing machinery in the host cells. However, the current tools have limitations concerning the identification of introns. This study proposes a new method, designated Fangorn Forest (F2), based on machine learning approaches to classify genera using an ab initio approach, i.e., using only the genomic sequence, as well as to predict and classify genes in the family Geminiviridae. In this investigation, nine genera of the family Geminiviridae and their related satellite DNAs were selected. We obtained two training sets, one for genus classification, containing attributes extracted from the complete genome of geminiviruses, while the other was made up to classify geminivirus genes, containing attributes extracted from ORFs taken from the complete genomes cited above. Three ML algorithms were applied on those datasets to build the predictive models: support vector machines, using the sequential minimal optimization training approach, random forest (RF), and multilayer perceptron. RF demonstrated a very high predictive power, achieving 0.966, 0.964, and 0.995 of precision, recall, and area under the curve (AUC), respectively, for genus classification. For gene classification, RF could reach 0.983, 0.983, and 0.998 of precision, recall, and AUC, respectively. Therefore, Fangorn Forest is proven to be an efficient method for classifying genera of the family Geminiviridae with high precision and effective gene prediction and classification. The method is freely accessible at www.geminivirus.org:8080/geminivirusdw/discoveryGeminivirus.jsp .

Applying Data Mining Techniques to Improve Breast Cancer Diagnosis.

PubMed

Diz, Joana; Marreiros, Goreti; Freitas, Alberto

2016-09-01

In the field of breast cancer research, and more than ever, new computer aided diagnosis based systems have been developed aiming to reduce diagnostic tests false-positives. Within this work, we present a data mining based approach which might support oncologists in the process of breast cancer classification and diagnosis. The present study aims to compare two breast cancer datasets and find the best methods in predicting benign/malignant lesions, breast density classification, and even for finding identification (mass / microcalcification distinction). To carry out these tasks, two matrices of texture features extraction were implemented using Matlab, and classified using data mining algorithms, on WEKA. Results revealed good percentages of accuracy for each class: 89.3 to 64.7 % - benign/malignant; 75.8 to 78.3 % - dense/fatty tissue; 71.0 to 83.1 % - finding identification. Among the different tests classifiers, Naive Bayes was the best to identify masses texture, and Random Forests was the first or second best classifier for the majority of tested groups.

Building rooftop classification using random forests for large-scale PV deployment

NASA Astrophysics Data System (ADS)

Assouline, Dan; Mohajeri, Nahid; Scartezzini, Jean-Louis

2017-10-01

Large scale solar Photovoltaic (PV) deployment on existing building rooftops has proven to be one of the most efficient and viable sources of renewable energy in urban areas. As it usually requires a potential analysis over the area of interest, a crucial step is to estimate the geometric characteristics of the building rooftops. In this paper, we introduce a multi-layer machine learning methodology to classify 6 roof types, 9 aspect (azimuth) classes and 5 slope (tilt) classes for all building rooftops in Switzerland, using GIS processing. We train Random Forests (RF), an ensemble learning algorithm, to build the classifiers. We use (2 × 2) [m2 ] LiDAR data (considering buildings and vegetation) to extract several rooftop features, and a generalised footprint polygon data to localize buildings. The roof classifier is trained and tested with 1252 labeled roofs from three different urban areas, namely Baden, Luzern, and Winterthur. The results for roof type classification show an average accuracy of 67%. The aspect and slope classifiers are trained and tested with 11449 labeled roofs in the Zurich periphery area. The results for aspect and slope classification show different accuracies depending on the classes: while some classes are well identified, other under-represented classes remain challenging to detect.

SU-C-207B-05: Tissue Segmentation of Computed Tomography Images Using a Random Forest Algorithm: A Feasibility Study

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

Polan, D; Brady, S; Kaufman, R

2016-06-15

Purpose: Develop an automated Random Forest algorithm for tissue segmentation of CT examinations. Methods: Seven materials were classified for segmentation: background, lung/internal gas, fat, muscle, solid organ parenchyma, blood/contrast, and bone using Matlab and the Trainable Weka Segmentation (TWS) plugin of FIJI. The following classifier feature filters of TWS were investigated: minimum, maximum, mean, and variance each evaluated over a pixel radius of 2n, (n = 0–4). Also noise reduction and edge preserving filters, Gaussian, bilateral, Kuwahara, and anisotropic diffusion, were evaluated. The algorithm used 200 trees with 2 features per node. A training data set was established using anmore » anonymized patient’s (male, 20 yr, 72 kg) chest-abdomen-pelvis CT examination. To establish segmentation ground truth, the training data were manually segmented using Eclipse planning software, and an intra-observer reproducibility test was conducted. Six additional patient data sets were segmented based on classifier data generated from the training data. Accuracy of segmentation was determined by calculating the Dice similarity coefficient (DSC) between manual and auto segmented images. Results: The optimized autosegmentation algorithm resulted in 16 features calculated using maximum, mean, variance, and Gaussian blur filters with kernel radii of 1, 2, and 4 pixels, in addition to the original CT number, and Kuwahara filter (linear kernel of 19 pixels). Ground truth had a DSC of 0.94 (range: 0.90–0.99) for adult and 0.92 (range: 0.85–0.99) for pediatric data sets across all seven segmentation classes. The automated algorithm produced segmentation with an average DSC of 0.85 ± 0.04 (range: 0.81–1.00) for the adult patients, and 0.86 ± 0.03 (range: 0.80–0.99) for the pediatric patients. Conclusion: The TWS Random Forest auto-segmentation algorithm was optimized for CT environment, and able to segment seven material classes over a range of body habitus and CT protocol parameters with an average DSC of 0.86 ± 0.04 (range: 0.80–0.99).« less

Sleep state classification using pressure sensor mats.

PubMed

Baran Pouyan, M; Nourani, M; Pompeo, M

2015-08-01

Sleep state detection is valuable in assessing patient's sleep quality and in-bed general behavior. In this paper, a novel classification approach of sleep states (sleep, pre-wake, wake) is proposed that uses only surface pressure sensors. In our method, a mobility metric is defined based on successive pressure body maps. Then, suitable statistical features are computed based on the mobility metric. Finally, a customized random forest classifier is employed to identify various classes including a new class for pre-wake state. Our algorithm achieves 96.1% and 88% accuracies for two (sleep, wake) and three (sleep, pre-wake, wake) class identification, respectively.

Evaluating data mining algorithms using molecular dynamics trajectories.

PubMed

Tatsis, Vasileios A; Tjortjis, Christos; Tzirakis, Panagiotis

2013-01-01

Molecular dynamics simulations provide a sample of a molecule's conformational space. Experiments on the mus time scale, resulting in large amounts of data, are nowadays routine. Data mining techniques such as classification provide a way to analyse such data. In this work, we evaluate and compare several classification algorithms using three data sets which resulted from computer simulations, of a potential enzyme mimetic biomolecule. We evaluated 65 classifiers available in the well-known data mining toolkit Weka, using 'classification' errors to assess algorithmic performance. Results suggest that: (i) 'meta' classifiers perform better than the other groups, when applied to molecular dynamics data sets; (ii) Random Forest and Rotation Forest are the best classifiers for all three data sets; and (iii) classification via clustering yields the highest classification error. Our findings are consistent with bibliographic evidence, suggesting a 'roadmap' for dealing with such data.

Learning-based 3T brain MRI segmentation with guidance from 7T MRI labeling.

PubMed

Deng, Minghui; Yu, Renping; Wang, Li; Shi, Feng; Yap, Pew-Thian; Shen, Dinggang

2016-12-01

Segmentation of brain magnetic resonance (MR) images into white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF) is crucial for brain structural measurement and disease diagnosis. Learning-based segmentation methods depend largely on the availability of good training ground truth. However, the commonly used 3T MR images are of insufficient image quality and often exhibit poor intensity contrast between WM, GM, and CSF. Therefore, they are not ideal for providing good ground truth label data for training learning-based methods. Recent advances in ultrahigh field 7T imaging make it possible to acquire images with excellent intensity contrast and signal-to-noise ratio. In this paper, the authors propose an algorithm based on random forest for segmenting 3T MR images by training a series of classifiers based on reliable labels obtained semiautomatically from 7T MR images. The proposed algorithm iteratively refines the probability maps of WM, GM, and CSF via a cascade of random forest classifiers for improved tissue segmentation. The proposed method was validated on two datasets, i.e., 10 subjects collected at their institution and 797 3T MR images from the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset. Specifically, for the mean Dice ratio of all 10 subjects, the proposed method achieved 94.52% ± 0.9%, 89.49% ± 1.83%, and 79.97% ± 4.32% for WM, GM, and CSF, respectively, which are significantly better than the state-of-the-art methods (p-values < 0.021). For the ADNI dataset, the group difference comparisons indicate that the proposed algorithm outperforms state-of-the-art segmentation methods. The authors have developed and validated a novel fully automated method for 3T brain MR image segmentation. © 2016 American Association of Physicists in Medicine.

Learning-based 3T brain MRI segmentation with guidance from 7T MRI labeling.

PubMed

Deng, Minghui; Yu, Renping; Wang, Li; Shi, Feng; Yap, Pew-Thian; Shen, Dinggang

2016-12-01

Segmentation of brain magnetic resonance (MR) images into white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF) is crucial for brain structural measurement and disease diagnosis. Learning-based segmentation methods depend largely on the availability of good training ground truth. However, the commonly used 3T MR images are of insufficient image quality and often exhibit poor intensity contrast between WM, GM, and CSF. Therefore, they are not ideal for providing good ground truth label data for training learning-based methods. Recent advances in ultrahigh field 7T imaging make it possible to acquire images with excellent intensity contrast and signal-to-noise ratio. In this paper, the authors propose an algorithm based on random forest for segmenting 3T MR images by training a series of classifiers based on reliable labels obtained semiautomatically from 7T MR images. The proposed algorithm iteratively refines the probability maps of WM, GM, and CSF via a cascade of random forest classifiers for improved tissue segmentation. The proposed method was validated on two datasets, i.e., 10 subjects collected at their institution and 797 3T MR images from the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset. Specifically, for the mean Dice ratio of all 10 subjects, the proposed method achieved 94.52% ± 0.9%, 89.49% ± 1.83%, and 79.97% ± 4.32% for WM, GM, and CSF, respectively, which are significantly better than the state-of-the-art methods (p-values < 0.021). For the ADNI dataset, the group difference comparisons indicate that the proposed algorithm outperforms state-of-the-art segmentation methods. The authors have developed and validated a novel fully automated method for 3T brain MR image segmentation.

Semantic segmentation of 3D textured meshes for urban scene analysis

NASA Astrophysics Data System (ADS)

Rouhani, Mohammad; Lafarge, Florent; Alliez, Pierre

2017-01-01

Classifying 3D measurement data has become a core problem in photogrammetry and 3D computer vision, since the rise of modern multiview geometry techniques, combined with affordable range sensors. We introduce a Markov Random Field-based approach for segmenting textured meshes generated via multi-view stereo into urban classes of interest. The input mesh is first partitioned into small clusters, referred to as superfacets, from which geometric and photometric features are computed. A random forest is then trained to predict the class of each superfacet as well as its similarity with the neighboring superfacets. Similarity is used to assign the weights of the Markov Random Field pairwise-potential and to account for contextual information between the classes. The experimental results illustrate the efficacy and accuracy of the proposed framework.

Methods for Real-Time Prediction of the Mode of Travel Using Smartphone-Based GPS and Accelerometer Data

PubMed Central

Martin, Bryan D.; Wolfson, Julian; Adomavicius, Gediminas; Fan, Yingling

2017-01-01

We propose and compare combinations of several methods for classifying transportation activity data from smartphone GPS and accelerometer sensors. We have two main objectives. First, we aim to classify our data as accurately as possible. Second, we aim to reduce the dimensionality of the data as much as possible in order to reduce the computational burden of the classification. We combine dimension reduction and classification algorithms and compare them with a metric that balances accuracy and dimensionality. In doing so, we develop a classification algorithm that accurately classifies five different modes of transportation (i.e., walking, biking, car, bus and rail) while being computationally simple enough to run on a typical smartphone. Further, we use data that required no behavioral changes from the smartphone users to collect. Our best classification model uses the random forest algorithm to achieve 96.8% accuracy. PMID:28885550

Methods for Real-Time Prediction of the Mode of Travel Using Smartphone-Based GPS and Accelerometer Data.

PubMed

Martin, Bryan D; Addona, Vittorio; Wolfson, Julian; Adomavicius, Gediminas; Fan, Yingling

2017-09-08

We propose and compare combinations of several methods for classifying transportation activity data from smartphone GPS and accelerometer sensors. We have two main objectives. First, we aim to classify our data as accurately as possible. Second, we aim to reduce the dimensionality of the data as much as possible in order to reduce the computational burden of the classification. We combine dimension reduction and classification algorithms and compare them with a metric that balances accuracy and dimensionality. In doing so, we develop a classification algorithm that accurately classifies five different modes of transportation (i.e., walking, biking, car, bus and rail) while being computationally simple enough to run on a typical smartphone. Further, we use data that required no behavioral changes from the smartphone users to collect. Our best classification model uses the random forest algorithm to achieve 96.8% accuracy.

Automated Classification of Consumer Health Information Needs in Patient Portal Messages.

PubMed

Cronin, Robert M; Fabbri, Daniel; Denny, Joshua C; Jackson, Gretchen Purcell

2015-01-01

Patients have diverse health information needs, and secure messaging through patient portals is an emerging means by which such needs are expressed and met. As patient portal adoption increases, growing volumes of secure messages may burden healthcare providers. Automated classification could expedite portal message triage and answering. We created four automated classifiers based on word content and natural language processing techniques to identify health information needs in 1000 patient-generated portal messages. Logistic regression and random forest classifiers detected single information needs well, with area under the curves of 0.804-0.914. A logistic regression classifier accurately found the set of needs within a message, with a Jaccard index of 0.859 (95% Confidence Interval: (0.847, 0.871)). Automated classification of consumer health information needs expressed in patient portal messages is feasible and may allow direct linking to relevant resources or creation of institutional resources for commonly expressed needs.

Automated Classification of Consumer Health Information Needs in Patient Portal Messages

PubMed Central

Cronin, Robert M.; Fabbri, Daniel; Denny, Joshua C.; Jackson, Gretchen Purcell

2015-01-01

Patients have diverse health information needs, and secure messaging through patient portals is an emerging means by which such needs are expressed and met. As patient portal adoption increases, growing volumes of secure messages may burden healthcare providers. Automated classification could expedite portal message triage and answering. We created four automated classifiers based on word content and natural language processing techniques to identify health information needs in 1000 patient-generated portal messages. Logistic regression and random forest classifiers detected single information needs well, with area under the curves of 0.804–0.914. A logistic regression classifier accurately found the set of needs within a message, with a Jaccard index of 0.859 (95% Confidence Interval: (0.847, 0.871)). Automated classification of consumer health information needs expressed in patient portal messages is feasible and may allow direct linking to relevant resources or creation of institutional resources for commonly expressed needs. PMID:26958285

Machine-learning-based classification of real-time tissue elastography for hepatic fibrosis in patients with chronic hepatitis B.

PubMed

Chen, Yang; Luo, Yan; Huang, Wei; Hu, Die; Zheng, Rong-Qin; Cong, Shu-Zhen; Meng, Fan-Kun; Yang, Hong; Lin, Hong-Jun; Sun, Yan; Wang, Xiu-Yan; Wu, Tao; Ren, Jie; Pei, Shu-Fang; Zheng, Ying; He, Yun; Hu, Yu; Yang, Na; Yan, Hongmei

2017-10-01

Hepatic fibrosis is a common middle stage of the pathological processes of chronic liver diseases. Clinical intervention during the early stages of hepatic fibrosis can slow the development of liver cirrhosis and reduce the risk of developing liver cancer. Performing a liver biopsy, the gold standard for viral liver disease management, has drawbacks such as invasiveness and a relatively high sampling error rate. Real-time tissue elastography (RTE), one of the most recently developed technologies, might be promising imaging technology because it is both noninvasive and provides accurate assessments of hepatic fibrosis. However, determining the stage of liver fibrosis from RTE images in a clinic is a challenging task. In this study, in contrast to the previous liver fibrosis index (LFI) method, which predicts the stage of diagnosis using RTE images and multiple regression analysis, we employed four classical classifiers (i.e., Support Vector Machine, Naïve Bayes, Random Forest and K-Nearest Neighbor) to build a decision-support system to improve the hepatitis B stage diagnosis performance. Eleven RTE image features were obtained from 513 subjects who underwent liver biopsies in this multicenter collaborative research. The experimental results showed that the adopted classifiers significantly outperformed the LFI method and that the Random Forest(RF) classifier provided the highest average accuracy among the four machine algorithms. This result suggests that sophisticated machine-learning methods can be powerful tools for evaluating the stage of hepatic fibrosis and show promise for clinical applications. Copyright © 2017 Elsevier Ltd. All rights reserved.

Mapping Forest Height in Gabon Using UAVSAR Multi-Baseline Polarimetric SAR Interferometry and Lidar Fusion

NASA Astrophysics Data System (ADS)

Simard, M.; Denbina, M. W.

2017-12-01

Using data collected by NASA's Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR) and Land, Vegetation, and Ice Sensor (LVIS) lidar, we have estimated forest canopy height for a number of study areas in the country of Gabon using a new machine learning data fusion approach. Using multi-baseline polarimetric synthetic aperture radar interferometry (PolInSAR) data collected by UAVSAR, forest heights can be estimated using the random volume over ground model. In the case of multi-baseline UAVSAR data consisting of many repeat passes with spatially separated flight tracks, we can estimate different forest height values for each different image pair, or baseline. In order to choose the best forest height estimate for each pixel, the baselines must be selected or ranked, taking care to avoid baselines with unsuitable spatial separation, or severe temporal decorrelation effects. The current baseline selection algorithms in the literature use basic quality metrics derived from the PolInSAR data which are not necessarily indicative of the true height accuracy in all cases. We have developed a new data fusion technique which treats PolInSAR baseline selection as a supervised classification problem, where the classifier is trained using a sparse sampling of lidar data within the PolInSAR coverage area. The classifier uses a large variety of PolInSAR-derived features as input, including radar backscatter as well as features based on the PolInSAR coherence region shape and the PolInSAR complex coherences. The resulting data fusion method produces forest height estimates which are more accurate than a purely radar-based approach, while having a larger coverage area than the input lidar training data, combining some of the strengths of each sensor. The technique demonstrates the strong potential for forest canopy height and above-ground biomass mapping using fusion of PolInSAR with data from future spaceborne lidar missions such as the upcoming Global Ecosystems Dynamics Investigation (GEDI) lidar.

Automatic detection of atrial fibrillation in cardiac vibration signals.

PubMed

Brueser, C; Diesel, J; Zink, M D H; Winter, S; Schauerte, P; Leonhardt, S

2013-01-01

We present a study on the feasibility of the automatic detection of atrial fibrillation (AF) from cardiac vibration signals (ballistocardiograms/BCGs) recorded by unobtrusive bedmounted sensors. The proposed system is intended as a screening and monitoring tool in home-healthcare applications and not as a replacement for ECG-based methods used in clinical environments. Based on BCG data recorded in a study with 10 AF patients, we evaluate and rank seven popular machine learning algorithms (naive Bayes, linear and quadratic discriminant analysis, support vector machines, random forests as well as bagged and boosted trees) for their performance in separating 30 s long BCG epochs into one of three classes: sinus rhythm, atrial fibrillation, and artifact. For each algorithm, feature subsets of a set of statistical time-frequency-domain and time-domain features were selected based on the mutual information between features and class labels as well as first- and second-order interactions among features. The classifiers were evaluated on a set of 856 epochs by means of 10-fold cross-validation. The best algorithm (random forests) achieved a Matthews correlation coefficient, mean sensitivity, and mean specificity of 0.921, 0.938, and 0.982, respectively.

High spatial resolution mapping of land cover types in a priority area for conservation in the Brazilian savanna

NASA Astrophysics Data System (ADS)

Ribeiro, F.; Roberts, D. A.; Hess, L. L.; Davis, F. W.; Caylor, K. K.; Nackoney, J.; Antunes Daldegan, G.

2017-12-01

Savannas are heterogeneous landscapes consisting of highly mixed land cover types that lack clear distinct boundaries. The Brazilian Cerrado is a Neotropical savanna considered a biodiversity hotspot for conservation due to its biodiversity richness and rapid transformation of its landscape by crop and pasture activities. The Cerrado is one of the most threatened Brazilian biomes and only 2.2% of its original extent is strictly protected. Accurate mapping and monitoring of its ecosystems and adjacent land use are important to select areas for conservation and to improve our understanding of the dynamics in this biome. Land cover mapping of savannas is difficult due to spectral similarity between land cover types resulting from similar vegetation structure, floristically similar components, generalization of land cover classes, and heterogeneity usually expressed as small patch sizes within the natural landscape. These factors are the major contributor to misclassification and low map accuracies among remote sensing studies in savannas. Specific challenges to map the Cerrado's land cover types are related to the spectral similarity between classes of land use and natural vegetation, such as natural grassland vs. cultivated pasture, and forest ecosystem vs. crops. This study seeks to classify and evaluate the land cover patterns across an area ranked as having extremely high priority for future conservation in the Cerrado. The main objective of this study is to identify the representativeness of each vegetation type across the landscape using high to moderate spatial resolution imagery using an automated scheme. A combination of pixel-based and object-based approaches were tested using RapidEye 3A imagery (5m spatial resolution) to classify the Cerrado's major land cover types. The random forest classifier was used to map the major ecosystems present across the area, and demonstrated to have an effective result with 68% of overall accuracy. Post-classification modification was performed to refine information to the major physiognomic groups of each ecosystem type. In this step, we used segmentation in eCognition, considering the random forest classification as input as well as other environmental layers (e.g. slope, soil types), which improved the overall classification to 75%.

Deep learning approach for classifying, detecting and predicting photometric redshifts of quasars in the Sloan Digital Sky Survey stripe 82

NASA Astrophysics Data System (ADS)

Pasquet-Itam, J.; Pasquet, J.

2018-04-01

We have applied a convolutional neural network (CNN) to classify and detect quasars in the Sloan Digital Sky Survey Stripe 82 and also to predict the photometric redshifts of quasars. The network takes the variability of objects into account by converting light curves into images. The width of the images, noted w, corresponds to the five magnitudes ugriz and the height of the images, noted h, represents the date of the observation. The CNN provides good results since its precision is 0.988 for a recall of 0.90, compared to a precision of 0.985 for the same recall with a random forest classifier. Moreover 175 new quasar candidates are found with the CNN considering a fixed recall of 0.97. The combination of probabilities given by the CNN and the random forest makes good performance even better with a precision of 0.99 for a recall of 0.90. For the redshift predictions, the CNN presents excellent results which are higher than those obtained with a feature extraction step and different classifiers (a K-nearest-neighbors, a support vector machine, a random forest and a Gaussian process classifier). Indeed, the accuracy of the CNN within |Δz| < 0.1 can reach 78.09%, within |Δz| < 0.2 reaches 86.15%, within |Δz| < 0.3 reaches 91.2% and the value of root mean square (rms) is 0.359. The performance of the KNN decreases for the three |Δz| regions, since within the accuracy of |Δz| < 0.1, |Δz| < 0.2, and |Δz| < 0.3 is 73.72%, 82.46%, and 90.09% respectively, and the value of rms amounts to 0.395. So the CNN successfully reduces the dispersion and the catastrophic redshifts of quasars. This new method is very promising for the future of big databases such as the Large Synoptic Survey Telescope. A table of the candidates is only available at the CDS via anonymous ftp to http://cdsarc.u-strasbg.fr (http://130.79.128.5) or via http://cdsarc.u-strasbg.fr/viz-bin/qcat?J/A+A/611/A97

Large-Scale Mixed Temperate Forest Mapping at the Single Tree Level using Airborne Laser Scanning

NASA Astrophysics Data System (ADS)

Scholl, V.; Morsdorf, F.; Ginzler, C.; Schaepman, M. E.

2017-12-01

Monitoring vegetation on a single tree level is critical to understand and model a variety of processes, functions, and changes in forest systems. Remote sensing technologies are increasingly utilized to complement and upscale the field-based measurements of forest inventories. Airborne laser scanning (ALS) systems provide valuable information in the vertical dimension for effective vegetation structure mapping. Although many algorithms exist to extract single tree segments from forest scans, they are often tuned to perform well in homogeneous coniferous or deciduous areas and are not successful in mixed forests. Other methods are too computationally expensive to apply operationally. The aim of this study was to develop a single tree detection workflow using leaf-off ALS data for the canton of Aargau in Switzerland. Aargau covers an area of over 1,400km2 and features mixed forests with various development stages and topography. Forest type was classified using random forests to guide local parameter selection. Canopy height model-based treetop maxima were detected and maintained based on the relationship between tree height and window size, used as a proxy to crown diameter. Watershed segmentation was used to generate crown polygons surrounding each maximum. The location, height, and crown dimensions of single trees were derived from the ALS returns within each polygon. Validation was performed through comparison with field measurements and extrapolated estimates from long-term monitoring plots of the Swiss National Forest Inventory within the framework of the Swiss Federal Institute for Forest, Snow, and Landscape Research. This method shows promise for robust, large-scale single tree detection in mixed forests. The single tree data will aid ecological studies as well as forest management practices. Figure description: Height-normalized ALS point cloud data (top) and resulting single tree segments (bottom) on the Laegeren mountain in Switzerland.

Application of Random Forests Methods to Diabetic Retinopathy Classification Analyses

PubMed Central

Casanova, Ramon; Saldana, Santiago; Chew, Emily Y.; Danis, Ronald P.; Greven, Craig M.; Ambrosius, Walter T.

2014-01-01

Background Diabetic retinopathy (DR) is one of the leading causes of blindness in the United States and world-wide. DR is a silent disease that may go unnoticed until it is too late for effective treatment. Therefore, early detection could improve the chances of therapeutic interventions that would alleviate its effects. Methodology Graded fundus photography and systemic data from 3443 ACCORD-Eye Study participants were used to estimate Random Forest (RF) and logistic regression classifiers. We studied the impact of sample size on classifier performance and the possibility of using RF generated class conditional probabilities as metrics describing DR risk. RF measures of variable importance are used to detect factors that affect classification performance. Principal Findings Both types of data were informative when discriminating participants with or without DR. RF based models produced much higher classification accuracy than those based on logistic regression. Combining both types of data did not increase accuracy but did increase statistical discrimination of healthy participants who subsequently did or did not have DR events during four years of follow-up. RF variable importance criteria revealed that microaneurysms counts in both eyes seemed to play the most important role in discrimination among the graded fundus variables, while the number of medicines and diabetes duration were the most relevant among the systemic variables. Conclusions and Significance We have introduced RF methods to DR classification analyses based on fundus photography data. In addition, we propose an approach to DR risk assessment based on metrics derived from graded fundus photography and systemic data. Our results suggest that RF methods could be a valuable tool to diagnose DR diagnosis and evaluate its progression. PMID:24940623

Water chemistry in 179 randomly selected Swedish headwater streams related to forest production, clear-felling and climate.

PubMed

Löfgren, Stefan; Fröberg, Mats; Yu, Jun; Nisell, Jakob; Ranneby, Bo

2014-12-01

From a policy perspective, it is important to understand forestry effects on surface waters from a landscape perspective. The EU Water Framework Directive demands remedial actions if not achieving good ecological status. In Sweden, 44 % of the surface water bodies have moderate ecological status or worse. Many of these drain catchments with a mosaic of managed forests. It is important for the forestry sector and water authorities to be able to identify where, in the forested landscape, special precautions are necessary. The aim of this study was to quantify the relations between forestry parameters and headwater stream concentrations of nutrients, organic matter and acid-base chemistry. The results are put into the context of regional climate, sulphur and nitrogen deposition, as well as marine influences. Water chemistry was measured in 179 randomly selected headwater streams from two regions in southwest and central Sweden, corresponding to 10 % of the Swedish land area. Forest status was determined from satellite images and Swedish National Forest Inventory data using the probabilistic classifier method, which was used to model stream water chemistry with Bayesian model averaging. The results indicate that concentrations of e.g. nitrogen, phosphorus and organic matter are related to factors associated with forest production but that it is not forestry per se that causes the excess losses. Instead, factors simultaneously affecting forest production and stream water chemistry, such as climate, extensive soil pools and nitrogen deposition, are the most likely candidates The relationships with clear-felled and wetland areas are likely to be direct effects.

Bayes classification of interferometric TOPSAR data

NASA Technical Reports Server (NTRS)

Michel, T. R.; Rodriguez, E.; Houshmand, B.; Carande, R.

1995-01-01

We report the Bayes classification of terrain types at different sites using airborne interferometric synthetic aperture radar (INSAR) data. A Gaussian maximum likelihood classifier was applied on multidimensional observations derived from the SAR intensity, the terrain elevation model, and the magnitude of the interferometric correlation. Training sets for forested, urban, agricultural, or bare areas were obtained either by selecting samples with known ground truth, or by k-means clustering of random sets of samples uniformly distributed across all sites, and subsequent assignments of these clusters using ground truth. The accuracy of the classifier was used to optimize the discriminating efficiency of the set of features that was chosen. The most important features include the SAR intensity, a canopy penetration depth model, and the terrain slope. We demonstrate the classifier's performance across sites using a unique set of training classes for the four main terrain categories. The scenes examined include San Francisco (CA) (predominantly urban and water), Mount Adams (WA) (forested with clear cuts), Pasadena (CA) (urban with mountains), and Antioch Hills (CA) (water, swamps, fields). Issues related to the effects of image calibration and the robustness of the classification to calibration errors are explored. The relative performance of single polarization Interferometric data classification is contrasted against classification schemes based on polarimetric SAR data.

Impact of professional foresters on timber harvests on West Virginia nonindustrial private forests

Treesearch

Stuart A. Moss; Eric Heitzman

2013-01-01

Timber harvests conducted on 90 nonindustrial private forest properties in West Virginia were investigated to determine the effects that professional foresters have on harvest and residual stand attributes. Harvests were classified based on the type of forester involved: (1) consulting/state service foresters representing landowners, (2) industry foresters representing...

Double-Windows-Based Motion Recognition in Multi-Floor Buildings Assisted by a Built-In Barometer.

PubMed

Liu, Maolin; Li, Huaiyu; Wang, Yuan; Li, Fei; Chen, Xiuwan

2018-04-01

Accelerometers, gyroscopes and magnetometers in smartphones are often used to recognize human motions. Since it is difficult to distinguish between vertical motions and horizontal motions in the data provided by these built-in sensors, the vertical motion recognition accuracy is relatively low. The emergence of a built-in barometer in smartphones improves the accuracy of motion recognition in the vertical direction. However, there is a lack of quantitative analysis and modelling of the barometer signals, which is the basis of barometer's application to motion recognition, and a problem of imbalanced data also exists. This work focuses on using the barometers inside smartphones for vertical motion recognition in multi-floor buildings through modelling and feature extraction of pressure signals. A novel double-windows pressure feature extraction method, which adopts two sliding time windows of different length, is proposed to balance recognition accuracy and response time. Then, a random forest classifier correlation rule is further designed to weaken the impact of imbalanced data on recognition accuracy. The results demonstrate that the recognition accuracy can reach 95.05% when pressure features and the improved random forest classifier are adopted. Specifically, the recognition accuracy of the stair and elevator motions is significantly improved with enhanced response time. The proposed approach proves effective and accurate, providing a robust strategy for increasing accuracy of vertical motions.

Improved predictive mapping of indoor radon concentrations using ensemble regression trees based on automatic clustering of geological units.

PubMed

Kropat, Georg; Bochud, Francois; Jaboyedoff, Michel; Laedermann, Jean-Pascal; Murith, Christophe; Palacios Gruson, Martha; Baechler, Sébastien

2015-09-01

According to estimations around 230 people die as a result of radon exposure in Switzerland. This public health concern makes reliable indoor radon prediction and mapping methods necessary in order to improve risk communication to the public. The aim of this study was to develop an automated method to classify lithological units according to their radon characteristics and to develop mapping and predictive tools in order to improve local radon prediction. About 240 000 indoor radon concentration (IRC) measurements in about 150 000 buildings were available for our analysis. The automated classification of lithological units was based on k-medoids clustering via pair-wise Kolmogorov distances between IRC distributions of lithological units. For IRC mapping and prediction we used random forests and Bayesian additive regression trees (BART). The automated classification groups lithological units well in terms of their IRC characteristics. Especially the IRC differences in metamorphic rocks like gneiss are well revealed by this method. The maps produced by random forests soundly represent the regional difference of IRCs in Switzerland and improve the spatial detail compared to existing approaches. We could explain 33% of the variations in IRC data with random forests. Additionally, the influence of a variable evaluated by random forests shows that building characteristics are less important predictors for IRCs than spatial/geological influences. BART could explain 29% of IRC variability and produced maps that indicate the prediction uncertainty. Ensemble regression trees are a powerful tool to model and understand the multidimensional influences on IRCs. Automatic clustering of lithological units complements this method by facilitating the interpretation of radon properties of rock types. This study provides an important element for radon risk communication. Future approaches should consider taking into account further variables like soil gas radon measurements as well as more detailed geological information. Copyright © 2015 Elsevier Ltd. All rights reserved.

Prediction of survival with multi-scale radiomic analysis in glioblastoma patients.

PubMed

Chaddad, Ahmad; Sabri, Siham; Niazi, Tamim; Abdulkarim, Bassam

2018-06-19

We propose a multiscale texture features based on Laplacian-of Gaussian (LoG) filter to predict progression free (PFS) and overall survival (OS) in patients newly diagnosed with glioblastoma (GBM). Experiments use the extracted features derived from 40 patients of GBM with T1-weighted imaging (T1-WI) and Fluid-attenuated inversion recovery (FLAIR) images that were segmented manually into areas of active tumor, necrosis, and edema. Multiscale texture features were extracted locally from each of these areas of interest using a LoG filter and the relation between features to OS and PFS was investigated using univariate (i.e., Spearman's rank correlation coefficient, log-rank test and Kaplan-Meier estimator) and multivariate analyses (i.e., Random Forest classifier). Three and seven features were statistically correlated with PFS and OS, respectively, with absolute correlation values between 0.32 and 0.36 and p < 0.05. Three features derived from active tumor regions only were associated with OS (p < 0.05) with hazard ratios (HR) of 2.9, 3, and 3.24, respectively. Combined features showed an AUC value of 85.37 and 85.54% for predicting the PFS and OS of GBM patients, respectively, using the random forest (RF) classifier. We presented a multiscale texture features to characterize the GBM regions and predict he PFS and OS. The efficiency achievable suggests that this technique can be developed into a GBM MR analysis system suitable for clinical use after a thorough validation involving more patients. Graphical abstract Scheme of the proposed model for characterizing the heterogeneity of GBM regions and predicting the overall survival and progression free survival of GBM patients. (1) Acquisition of pretreatment MRI images; (2) Affine registration of T1-WI image with its corresponding FLAIR images, and GBM subtype (phenotypes) labelling; (3) Extraction of nine texture features from the three texture scales fine, medium, and coarse derived from each of GBM regions; (4) Comparing heterogeneity between GBM regions by ANOVA test; Survival analysis using Univariate (Spearman rank correlation between features and survival (i.e., PFS and OS) based on each of the GBM regions, Kaplan-Meier estimator and log-rank test to predict the PFS and OS of patient groups that grouped based on median of feature), and multivariate (random forest model) for predicting the PFS and OS of patients groups that grouped based on median of PFS and OS.

Modeling Verdict Outcomes Using Social Network Measures: The Watergate and Caviar Network Cases

PubMed Central

2016-01-01

Modelling criminal trial verdict outcomes using social network measures is an emerging research area in quantitative criminology. Few studies have yet analyzed which of these measures are the most important for verdict modelling or which data classification techniques perform best for this application. To compare the performance of different techniques in classifying members of a criminal network, this article applies three different machine learning classifiers–Logistic Regression, Naïve Bayes and Random Forest–with a range of social network measures and the necessary databases to model the verdicts in two real–world cases: the U.S. Watergate Conspiracy of the 1970’s and the now–defunct Canada–based international drug trafficking ring known as the Caviar Network. In both cases it was found that the Random Forest classifier did better than either Logistic Regression or Naïve Bayes, and its superior performance was statistically significant. This being so, Random Forest was used not only for classification but also to assess the importance of the measures. For the Watergate case, the most important one proved to be betweenness centrality while for the Caviar Network, it was the effective size of the network. These results are significant because they show that an approach combining machine learning with social network analysis not only can generate accurate classification models but also helps quantify the importance social network variables in modelling verdict outcomes. We conclude our analysis with a discussion and some suggestions for future work in verdict modelling using social network measures. PMID:26824351

Comparison of Different Features and Classifiers for Driver Fatigue Detection Based on a Single EEG Channel

PubMed Central

2017-01-01

Driver fatigue has become an important factor to traffic accidents worldwide, and effective detection of driver fatigue has major significance for public health. The purpose method employs entropy measures for feature extraction from a single electroencephalogram (EEG) channel. Four types of entropies measures, sample entropy (SE), fuzzy entropy (FE), approximate entropy (AE), and spectral entropy (PE), were deployed for the analysis of original EEG signal and compared by ten state-of-the-art classifiers. Results indicate that optimal performance of single channel is achieved using a combination of channel CP4, feature FE, and classifier Random Forest (RF). The highest accuracy can be up to 96.6%, which has been able to meet the needs of real applications. The best combination of channel + features + classifier is subject-specific. In this work, the accuracy of FE as the feature is far greater than the Acc of other features. The accuracy using classifier RF is the best, while that of classifier SVM with linear kernel is the worst. The impact of channel selection on the Acc is larger. The performance of various channels is very different. PMID:28255330

Comparison of Different Features and Classifiers for Driver Fatigue Detection Based on a Single EEG Channel.

PubMed

Hu, Jianfeng

2017-01-01

Driver fatigue has become an important factor to traffic accidents worldwide, and effective detection of driver fatigue has major significance for public health. The purpose method employs entropy measures for feature extraction from a single electroencephalogram (EEG) channel. Four types of entropies measures, sample entropy (SE), fuzzy entropy (FE), approximate entropy (AE), and spectral entropy (PE), were deployed for the analysis of original EEG signal and compared by ten state-of-the-art classifiers. Results indicate that optimal performance of single channel is achieved using a combination of channel CP4, feature FE, and classifier Random Forest (RF). The highest accuracy can be up to 96.6%, which has been able to meet the needs of real applications. The best combination of channel + features + classifier is subject-specific. In this work, the accuracy of FE as the feature is far greater than the Acc of other features. The accuracy using classifier RF is the best, while that of classifier SVM with linear kernel is the worst. The impact of channel selection on the Acc is larger. The performance of various channels is very different.

Linear Subpixel Learning Algorithm for Land Cover Classification from WELD using High Performance Computing

NASA Technical Reports Server (NTRS)

Kumar, Uttam; Nemani, Ramakrishna R.; Ganguly, Sangram; Kalia, Subodh; Michaelis, Andrew

2017-01-01

In this work, we use a Fully Constrained Least Squares Subpixel Learning Algorithm to unmix global WELD (Web Enabled Landsat Data) to obtain fractions or abundances of substrate (S), vegetation (V) and dark objects (D) classes. Because of the sheer nature of data and compute needs, we leveraged the NASA Earth Exchange (NEX) high performance computing architecture to optimize and scale our algorithm for large-scale processing. Subsequently, the S-V-D abundance maps were characterized into 4 classes namely, forest, farmland, water and urban areas (with NPP-VIIRS-national polar orbiting partnership visible infrared imaging radiometer suite nighttime lights data) over California, USA using Random Forest classifier. Validation of these land cover maps with NLCD (National Land Cover Database) 2011 products and NAFD (North American Forest Dynamics) static forest cover maps showed that an overall classification accuracy of over 91 percent was achieved, which is a 6 percent improvement in unmixing based classification relative to per-pixel-based classification. As such, abundance maps continue to offer an useful alternative to high-spatial resolution data derived classification maps for forest inventory analysis, multi-class mapping for eco-climatic models and applications, fast multi-temporal trend analysis and for societal and policy-relevant applications needed at the watershed scale.

Linear Subpixel Learning Algorithm for Land Cover Classification from WELD using High Performance Computing

NASA Astrophysics Data System (ADS)

Ganguly, S.; Kumar, U.; Nemani, R. R.; Kalia, S.; Michaelis, A.

2017-12-01

In this work, we use a Fully Constrained Least Squares Subpixel Learning Algorithm to unmix global WELD (Web Enabled Landsat Data) to obtain fractions or abundances of substrate (S), vegetation (V) and dark objects (D) classes. Because of the sheer nature of data and compute needs, we leveraged the NASA Earth Exchange (NEX) high performance computing architecture to optimize and scale our algorithm for large-scale processing. Subsequently, the S-V-D abundance maps were characterized into 4 classes namely, forest, farmland, water and urban areas (with NPP-VIIRS - national polar orbiting partnership visible infrared imaging radiometer suite nighttime lights data) over California, USA using Random Forest classifier. Validation of these land cover maps with NLCD (National Land Cover Database) 2011 products and NAFD (North American Forest Dynamics) static forest cover maps showed that an overall classification accuracy of over 91% was achieved, which is a 6% improvement in unmixing based classification relative to per-pixel based classification. As such, abundance maps continue to offer an useful alternative to high-spatial resolution data derived classification maps for forest inventory analysis, multi-class mapping for eco-climatic models and applications, fast multi-temporal trend analysis and for societal and policy-relevant applications needed at the watershed scale.

Evaluating the Effectiveness of Flood Control Strategies in Contrasting Urban Watersheds and Implications for Houston's Future Flood Vulnerability

NASA Astrophysics Data System (ADS)

Ganguly, S.; Kumar, U.; Nemani, R. R.; Kalia, S.; Michaelis, A.

2016-12-01

In this work, we use a Fully Constrained Least Squares Subpixel Learning Algorithm to unmix global WELD (Web Enabled Landsat Data) to obtain fractions or abundances of substrate (S), vegetation (V) and dark objects (D) classes. Because of the sheer nature of data and compute needs, we leveraged the NASA Earth Exchange (NEX) high performance computing architecture to optimize and scale our algorithm for large-scale processing. Subsequently, the S-V-D abundance maps were characterized into 4 classes namely, forest, farmland, water and urban areas (with NPP-VIIRS - national polar orbiting partnership visible infrared imaging radiometer suite nighttime lights data) over California, USA using Random Forest classifier. Validation of these land cover maps with NLCD (National Land Cover Database) 2011 products and NAFD (North American Forest Dynamics) static forest cover maps showed that an overall classification accuracy of over 91% was achieved, which is a 6% improvement in unmixing based classification relative to per-pixel based classification. As such, abundance maps continue to offer an useful alternative to high-spatial resolution data derived classification maps for forest inventory analysis, multi-class mapping for eco-climatic models and applications, fast multi-temporal trend analysis and for societal and policy-relevant applications needed at the watershed scale.

Quantification of the heterogeneity of prognostic cellular biomarkers in ewing sarcoma using automated image and random survival forest analysis.

PubMed

Bühnemann, Claudia; Li, Simon; Yu, Haiyue; Branford White, Harriet; Schäfer, Karl L; Llombart-Bosch, Antonio; Machado, Isidro; Picci, Piero; Hogendoorn, Pancras C W; Athanasou, Nicholas A; Noble, J Alison; Hassan, A Bassim

2014-01-01

Driven by genomic somatic variation, tumour tissues are typically heterogeneous, yet unbiased quantitative methods are rarely used to analyse heterogeneity at the protein level. Motivated by this problem, we developed automated image segmentation of images of multiple biomarkers in Ewing sarcoma to generate distributions of biomarkers between and within tumour cells. We further integrate high dimensional data with patient clinical outcomes utilising random survival forest (RSF) machine learning. Using material from cohorts of genetically diagnosed Ewing sarcoma with EWSR1 chromosomal translocations, confocal images of tissue microarrays were segmented with level sets and watershed algorithms. Each cell nucleus and cytoplasm were identified in relation to DAPI and CD99, respectively, and protein biomarkers (e.g. Ki67, pS6, Foxo3a, EGR1, MAPK) localised relative to nuclear and cytoplasmic regions of each cell in order to generate image feature distributions. The image distribution features were analysed with RSF in relation to known overall patient survival from three separate cohorts (185 informative cases). Variation in pre-analytical processing resulted in elimination of a high number of non-informative images that had poor DAPI localisation or biomarker preservation (67 cases, 36%). The distribution of image features for biomarkers in the remaining high quality material (118 cases, 104 features per case) were analysed by RSF with feature selection, and performance assessed using internal cross-validation, rather than a separate validation cohort. A prognostic classifier for Ewing sarcoma with low cross-validation error rates (0.36) was comprised of multiple features, including the Ki67 proliferative marker and a sub-population of cells with low cytoplasmic/nuclear ratio of CD99. Through elimination of bias, the evaluation of high-dimensionality biomarker distribution within cell populations of a tumour using random forest analysis in quality controlled tumour material could be achieved. Such an automated and integrated methodology has potential application in the identification of prognostic classifiers based on tumour cell heterogeneity.

Predicting diabetes mellitus using SMOTE and ensemble machine learning approach: The Henry Ford ExercIse Testing (FIT) project.

PubMed

Alghamdi, Manal; Al-Mallah, Mouaz; Keteyian, Steven; Brawner, Clinton; Ehrman, Jonathan; Sakr, Sherif

2017-01-01

Machine learning is becoming a popular and important approach in the field of medical research. In this study, we investigate the relative performance of various machine learning methods such as Decision Tree, Naïve Bayes, Logistic Regression, Logistic Model Tree and Random Forests for predicting incident diabetes using medical records of cardiorespiratory fitness. In addition, we apply different techniques to uncover potential predictors of diabetes. This FIT project study used data of 32,555 patients who are free of any known coronary artery disease or heart failure who underwent clinician-referred exercise treadmill stress testing at Henry Ford Health Systems between 1991 and 2009 and had a complete 5-year follow-up. At the completion of the fifth year, 5,099 of those patients have developed diabetes. The dataset contained 62 attributes classified into four categories: demographic characteristics, disease history, medication use history, and stress test vital signs. We developed an Ensembling-based predictive model using 13 attributes that were selected based on their clinical importance, Multiple Linear Regression, and Information Gain Ranking methods. The negative effect of the imbalance class of the constructed model was handled by Synthetic Minority Oversampling Technique (SMOTE). The overall performance of the predictive model classifier was improved by the Ensemble machine learning approach using the Vote method with three Decision Trees (Naïve Bayes Tree, Random Forest, and Logistic Model Tree) and achieved high accuracy of prediction (AUC = 0.92). The study shows the potential of ensembling and SMOTE approaches for predicting incident diabetes using cardiorespiratory fitness data.

Sensing Urban Land-Use Patterns by Integrating Google Tensorflow and Scene-Classification Models

NASA Astrophysics Data System (ADS)

Yao, Y.; Liang, H.; Li, X.; Zhang, J.; He, J.

2017-09-01

With the rapid progress of China's urbanization, research on the automatic detection of land-use patterns in Chinese cities is of substantial importance. Deep learning is an effective method to extract image features. To take advantage of the deep-learning method in detecting urban land-use patterns, we applied a transfer-learning-based remote-sensing image approach to extract and classify features. Using the Google Tensorflow framework, a powerful convolution neural network (CNN) library was created. First, the transferred model was previously trained on ImageNet, one of the largest object-image data sets, to fully develop the model's ability to generate feature vectors of standard remote-sensing land-cover data sets (UC Merced and WHU-SIRI). Then, a random-forest-based classifier was constructed and trained on these generated vectors to classify the actual urban land-use pattern on the scale of traffic analysis zones (TAZs). To avoid the multi-scale effect of remote-sensing imagery, a large random patch (LRP) method was used. The proposed method could efficiently obtain acceptable accuracy (OA = 0.794, Kappa = 0.737) for the study area. In addition, the results show that the proposed method can effectively overcome the multi-scale effect that occurs in urban land-use classification at the irregular land-parcel level. The proposed method can help planners monitor dynamic urban land use and evaluate the impact of urban-planning schemes.

Dynamic species classification of microorganisms across time, abiotic and biotic environments—A sliding window approach

PubMed Central

Griffiths, Jason I.; Fronhofer, Emanuel A.; Garnier, Aurélie; Seymour, Mathew; Altermatt, Florian; Petchey, Owen L.

2017-01-01

The development of video-based monitoring methods allows for rapid, dynamic and accurate monitoring of individuals or communities, compared to slower traditional methods, with far reaching ecological and evolutionary applications. Large amounts of data are generated using video-based methods, which can be effectively processed using machine learning (ML) algorithms into meaningful ecological information. ML uses user defined classes (e.g. species), derived from a subset (i.e. training data) of video-observed quantitative features (e.g. phenotypic variation), to infer classes in subsequent observations. However, phenotypic variation often changes due to environmental conditions, which may lead to poor classification, if environmentally induced variation in phenotypes is not accounted for. Here we describe a framework for classifying species under changing environmental conditions based on the random forest classification. A sliding window approach was developed that restricts temporal and environmentally conditions to improve the classification. We tested our approach by applying the classification framework to experimental data. The experiment used a set of six ciliate species to monitor changes in community structure and behavior over hundreds of generations, in dozens of species combinations and across a temperature gradient. Differences in biotic and abiotic conditions caused simplistic classification approaches to be unsuccessful. In contrast, the sliding window approach allowed classification to be highly successful, as phenotypic differences driven by environmental change, could be captured by the classifier. Importantly, classification using the random forest algorithm showed comparable success when validated against traditional, slower, manual identification. Our framework allows for reliable classification in dynamic environments, and may help to improve strategies for long-term monitoring of species in changing environments. Our classification pipeline can be applied in fields assessing species community dynamics, such as eco-toxicology, ecology and evolutionary ecology. PMID:28472193

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

Ukwatta, T. N.; Wozniak, P. R.; Gehrels, N.

Studies of high-redshift gamma-ray bursts (GRBs) provide important information about the early Universe such as the rates of stellar collapsars and mergers, the metallicity content, constraints on the re-ionization period, and probes of the Hubble expansion. Rapid selection of high-z candidates from GRB samples reported in real time by dedicated space missions such as Swift is the key to identifying the most distant bursts before the optical afterglow becomes too dim to warrant a good spectrum. Here, we introduce ‘machine-z’, a redshift prediction algorithm and a ‘high-z’ classifier for Swift GRBs based on machine learning. Our method relies exclusively onmore » canonical data commonly available within the first few hours after the GRB trigger. Using a sample of 284 bursts with measured redshifts, we trained a randomized ensemble of decision trees (random forest) to perform both regression and classification. Cross-validated performance studies show that the correlation coefficient between machine-z predictions and the true redshift is nearly 0.6. At the same time, our high-z classifier can achieve 80 per cent recall of true high-redshift bursts, while incurring a false positive rate of 20 per cent. With 40 per cent false positive rate the classifier can achieve ~100 per cent recall. As a result, the most reliable selection of high-redshift GRBs is obtained by combining predictions from both the high-z classifier and the machine-z regressor.« less

Machine-z: Rapid Machine-Learned Redshift Indicator for Swift Gamma-Ray Bursts

NASA Technical Reports Server (NTRS)

Ukwatta, T. N.; Wozniak, P. R.; Gehrels, N.

2016-01-01

Studies of high-redshift gamma-ray bursts (GRBs) provide important information about the early Universe such as the rates of stellar collapsars and mergers, the metallicity content, constraints on the re-ionization period, and probes of the Hubble expansion. Rapid selection of high-z candidates from GRB samples reported in real time by dedicated space missions such as Swift is the key to identifying the most distant bursts before the optical afterglow becomes too dim to warrant a good spectrum. Here, we introduce 'machine-z', a redshift prediction algorithm and a 'high-z' classifier for Swift GRBs based on machine learning. Our method relies exclusively on canonical data commonly available within the first few hours after the GRB trigger. Using a sample of 284 bursts with measured redshifts, we trained a randomized ensemble of decision trees (random forest) to perform both regression and classification. Cross-validated performance studies show that the correlation coefficient between machine-z predictions and the true redshift is nearly 0.6. At the same time, our high-z classifier can achieve 80 per cent recall of true high-redshift bursts, while incurring a false positive rate of 20 per cent. With 40 per cent false positive rate the classifier can achieve approximately 100 per cent recall. The most reliable selection of high-redshift GRBs is obtained by combining predictions from both the high-z classifier and the machine-z regressor.

Bilayer segmentation of webcam videos using tree-based classifiers.

PubMed

Yin, Pei; Criminisi, Antonio; Winn, John; Essa, Irfan

2011-01-01

This paper presents an automatic segmentation algorithm for video frames captured by a (monocular) webcam that closely approximates depth segmentation from a stereo camera. The frames are segmented into foreground and background layers that comprise a subject (participant) and other objects and individuals. The algorithm produces correct segmentations even in the presence of large background motion with a nearly stationary foreground. This research makes three key contributions: First, we introduce a novel motion representation, referred to as "motons," inspired by research in object recognition. Second, we propose estimating the segmentation likelihood from the spatial context of motion. The estimation is efficiently learned by random forests. Third, we introduce a general taxonomy of tree-based classifiers that facilitates both theoretical and experimental comparisons of several known classification algorithms and generates new ones. In our bilayer segmentation algorithm, diverse visual cues such as motion, motion context, color, contrast, and spatial priors are fused by means of a conditional random field (CRF) model. Segmentation is then achieved by binary min-cut. Experiments on many sequences of our videochat application demonstrate that our algorithm, which requires no initialization, is effective in a variety of scenes, and the segmentation results are comparable to those obtained by stereo systems.

A tale of two "forests": random forest machine learning AIDS tropical forest carbon mapping.

PubMed

Mascaro, Joseph; Asner, Gregory P; Knapp, David E; Kennedy-Bowdoin, Ty; Martin, Roberta E; Anderson, Christopher; Higgins, Mark; Chadwick, K Dana

2014-01-01

Accurate and spatially-explicit maps of tropical forest carbon stocks are needed to implement carbon offset mechanisms such as REDD+ (Reduced Deforestation and Degradation Plus). The Random Forest machine learning algorithm may aid carbon mapping applications using remotely-sensed data. However, Random Forest has never been compared to traditional and potentially more reliable techniques such as regionally stratified sampling and upscaling, and it has rarely been employed with spatial data. Here, we evaluated the performance of Random Forest in upscaling airborne LiDAR (Light Detection and Ranging)-based carbon estimates compared to the stratification approach over a 16-million hectare focal area of the Western Amazon. We considered two runs of Random Forest, both with and without spatial contextual modeling by including--in the latter case--x, and y position directly in the model. In each case, we set aside 8 million hectares (i.e., half of the focal area) for validation; this rigorous test of Random Forest went above and beyond the internal validation normally compiled by the algorithm (i.e., called "out-of-bag"), which proved insufficient for this spatial application. In this heterogeneous region of Northern Peru, the model with spatial context was the best preforming run of Random Forest, and explained 59% of LiDAR-based carbon estimates within the validation area, compared to 37% for stratification or 43% by Random Forest without spatial context. With the 60% improvement in explained variation, RMSE against validation LiDAR samples improved from 33 to 26 Mg C ha(-1) when using Random Forest with spatial context. Our results suggest that spatial context should be considered when using Random Forest, and that doing so may result in substantially improved carbon stock modeling for purposes of climate change mitigation.

Developing a radiomics framework for classifying non-small cell lung carcinoma subtypes

NASA Astrophysics Data System (ADS)

Yu, Dongdong; Zang, Yali; Dong, Di; Zhou, Mu; Gevaert, Olivier; Fang, Mengjie; Shi, Jingyun; Tian, Jie

2017-03-01

Patient-targeted treatment of non-small cell lung carcinoma (NSCLC) has been well documented according to the histologic subtypes over the past decade. In parallel, recent development of quantitative image biomarkers has recently been highlighted as important diagnostic tools to facilitate histological subtype classification. In this study, we present a radiomics analysis that classifies the adenocarcinoma (ADC) and squamous cell carcinoma (SqCC). We extract 52-dimensional, CT-based features (7 statistical features and 45 image texture features) to represent each nodule. We evaluate our approach on a clinical dataset including 324 ADCs and 110 SqCCs patients with CT image scans. Classification of these features is performed with four different machine-learning classifiers including Support Vector Machines with Radial Basis Function kernel (RBF-SVM), Random forest (RF), K-nearest neighbor (KNN), and RUSBoost algorithms. To improve the classifiers' performance, optimal feature subset is selected from the original feature set by using an iterative forward inclusion and backward eliminating algorithm. Extensive experimental results demonstrate that radiomics features achieve encouraging classification results on both complete feature set (AUC=0.89) and optimal feature subset (AUC=0.91).

Semantic classification of urban buildings combining VHR image and GIS data: An improved random forest approach

NASA Astrophysics Data System (ADS)

Du, Shihong; Zhang, Fangli; Zhang, Xiuyuan

2015-07-01

While most existing studies have focused on extracting geometric information on buildings, only a few have concentrated on semantic information. The lack of semantic information cannot satisfy many demands on resolving environmental and social issues. This study presents an approach to semantically classify buildings into much finer categories than those of existing studies by learning random forest (RF) classifier from a large number of imbalanced samples with high-dimensional features. First, a two-level segmentation mechanism combining GIS and VHR image produces single image objects at a large scale and intra-object components at a small scale. Second, a semi-supervised method chooses a large number of unbiased samples by considering the spatial proximity and intra-cluster similarity of buildings. Third, two important improvements in RF classifier are made: a voting-distribution ranked rule for reducing the influences of imbalanced samples on classification accuracy and a feature importance measurement for evaluating each feature's contribution to the recognition of each category. Fourth, the semantic classification of urban buildings is practically conducted in Beijing city, and the results demonstrate that the proposed approach is effective and accurate. The seven categories used in the study are finer than those in existing work and more helpful to studying many environmental and social problems.

Automatic Classification of Time-variable X-Ray Sources

NASA Astrophysics Data System (ADS)

Lo, Kitty K.; Farrell, Sean; Murphy, Tara; Gaensler, B. M.

2014-05-01

To maximize the discovery potential of future synoptic surveys, especially in the field of transient science, it will be necessary to use automatic classification to identify some of the astronomical sources. The data mining technique of supervised classification is suitable for this problem. Here, we present a supervised learning method to automatically classify variable X-ray sources in the Second XMM-Newton Serendipitous Source Catalog (2XMMi-DR2). Random Forest is our classifier of choice since it is one of the most accurate learning algorithms available. Our training set consists of 873 variable sources and their features are derived from time series, spectra, and other multi-wavelength contextual information. The 10 fold cross validation accuracy of the training data is ~97% on a 7 class data set. We applied the trained classification model to 411 unknown variable 2XMM sources to produce a probabilistically classified catalog. Using the classification margin and the Random Forest derived outlier measure, we identified 12 anomalous sources, of which 2XMM J180658.7-500250 appears to be the most unusual source in the sample. Its X-ray spectra is suggestive of a ultraluminous X-ray source but its variability makes it highly unusual. Machine-learned classification and anomaly detection will facilitate scientific discoveries in the era of all-sky surveys.

Analysis of the 1996 Wisconsin forest statistics by habitat type.

Treesearch

John Kotar; Joseph A. Kovach; Gary Brand

1999-01-01

The fifth inventory of Wisconsin's forests is presented from the perspective of habitat type as a classification tool. Habitat type classifies forests based on the species composition of the understory plant community. Various forest attributes are summarized by habitat type and management implications are discussed.

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.

Computer-aided detection of prostate cancer in T2-weighted MRI within the peripheral zone

NASA Astrophysics Data System (ADS)

Rampun, Andrik; Zheng, Ling; Malcolm, Paul; Tiddeman, Bernie; Zwiggelaar, Reyer

2016-07-01

In this paper we propose a prostate cancer computer-aided diagnosis (CAD) system and suggest a set of discriminant texture descriptors extracted from T2-weighted MRI data which can be used as a good basis for a multimodality system. For this purpose, 215 texture descriptors were extracted and eleven different classifiers were employed to achieve the best possible results. The proposed method was tested based on 418 T2-weighted MR images taken from 45 patients and evaluated using 9-fold cross validation with five patients in each fold. The results demonstrated comparable results to existing CAD systems using multimodality MRI. We achieved an area under the receiver operating curve (A z ) values equal to 90.0%+/- 7.6% , 89.5%+/- 8.9% , 87.9%+/- 9.3% and 87.4%+/- 9.2% for Bayesian networks, ADTree, random forest and multilayer perceptron classifiers, respectively, while a meta-voting classifier using average probability as a combination rule achieved 92.7%+/- 7.4% .

Learning ensemble classifiers for diabetic retinopathy assessment.

PubMed

Saleh, Emran; Błaszczyński, Jerzy; Moreno, Antonio; Valls, Aida; Romero-Aroca, Pedro; de la Riva-Fernández, Sofia; Słowiński, Roman

2018-04-01

Diabetic retinopathy is one of the most common comorbidities of diabetes. Unfortunately, the recommended annual screening of the eye fundus of diabetic patients is too resource-consuming. Therefore, it is necessary to develop tools that may help doctors to determine the risk of each patient to attain this condition, so that patients with a low risk may be screened less frequently and the use of resources can be improved. This paper explores the use of two kinds of ensemble classifiers learned from data: fuzzy random forest and dominance-based rough set balanced rule ensemble. These classifiers use a small set of attributes which represent main risk factors to determine whether a patient is in risk of developing diabetic retinopathy. The levels of specificity and sensitivity obtained in the presented study are over 80%. This study is thus a first successful step towards the construction of a personalized decision support system that could help physicians in daily clinical practice. Copyright © 2017 Elsevier B.V. All rights reserved.

A Noise-Assisted Data Analysis Method for Automatic EOG-Based Sleep Stage Classification Using Ensemble Learning.

PubMed

Olesen, Alexander Neergaard; Christensen, Julie A E; Sorensen, Helge B D; Jennum, Poul J

2016-08-01

Reducing the number of recording modalities for sleep staging research can benefit both researchers and patients, under the condition that they provide as accurate results as conventional systems. This paper investigates the possibility of exploiting the multisource nature of the electrooculography (EOG) signals by presenting a method for automatic sleep staging using the complete ensemble empirical mode decomposition with adaptive noise algorithm, and a random forest classifier. It achieves a high overall accuracy of 82% and a Cohen's kappa of 0.74 indicating substantial agreement between automatic and manual scoring.

A Comparison of Physiological Signal Analysis Techniques and Classifiers for Automatic Emotional Evaluation of Audiovisual Contents

PubMed Central

Colomer Granero, Adrián; Fuentes-Hurtado, Félix; Naranjo Ornedo, Valery; Guixeres Provinciale, Jaime; Ausín, Jose M.; Alcañiz Raya, Mariano

2016-01-01

This work focuses on finding the most discriminatory or representative features that allow to classify commercials according to negative, neutral and positive effectiveness based on the Ace Score index. For this purpose, an experiment involving forty-seven participants was carried out. In this experiment electroencephalography (EEG), electrocardiography (ECG), Galvanic Skin Response (GSR) and respiration data were acquired while subjects were watching a 30-min audiovisual content. This content was composed by a submarine documentary and nine commercials (one of them the ad under evaluation). After the signal pre-processing, four sets of features were extracted from the physiological signals using different state-of-the-art metrics. These features computed in time and frequency domains are the inputs to several basic and advanced classifiers. An average of 89.76% of the instances was correctly classified according to the Ace Score index. The best results were obtained by a classifier consisting of a combination between AdaBoost and Random Forest with automatic selection of features. The selected features were those extracted from GSR and HRV signals. These results are promising in the audiovisual content evaluation field by means of physiological signal processing. PMID:27471462

A Comparison of Physiological Signal Analysis Techniques and Classifiers for Automatic Emotional Evaluation of Audiovisual Contents.

PubMed

Colomer Granero, Adrián; Fuentes-Hurtado, Félix; Naranjo Ornedo, Valery; Guixeres Provinciale, Jaime; Ausín, Jose M; Alcañiz Raya, Mariano

2016-01-01

This work focuses on finding the most discriminatory or representative features that allow to classify commercials according to negative, neutral and positive effectiveness based on the Ace Score index. For this purpose, an experiment involving forty-seven participants was carried out. In this experiment electroencephalography (EEG), electrocardiography (ECG), Galvanic Skin Response (GSR) and respiration data were acquired while subjects were watching a 30-min audiovisual content. This content was composed by a submarine documentary and nine commercials (one of them the ad under evaluation). After the signal pre-processing, four sets of features were extracted from the physiological signals using different state-of-the-art metrics. These features computed in time and frequency domains are the inputs to several basic and advanced classifiers. An average of 89.76% of the instances was correctly classified according to the Ace Score index. The best results were obtained by a classifier consisting of a combination between AdaBoost and Random Forest with automatic selection of features. The selected features were those extracted from GSR and HRV signals. These results are promising in the audiovisual content evaluation field by means of physiological signal processing.

A Tale of Two “Forests”: Random Forest Machine Learning Aids Tropical Forest Carbon Mapping

PubMed Central

Mascaro, Joseph; Asner, Gregory P.; Knapp, David E.; Kennedy-Bowdoin, Ty; Martin, Roberta E.; Anderson, Christopher; Higgins, Mark; Chadwick, K. Dana

2014-01-01

Accurate and spatially-explicit maps of tropical forest carbon stocks are needed to implement carbon offset mechanisms such as REDD+ (Reduced Deforestation and Degradation Plus). The Random Forest machine learning algorithm may aid carbon mapping applications using remotely-sensed data. However, Random Forest has never been compared to traditional and potentially more reliable techniques such as regionally stratified sampling and upscaling, and it has rarely been employed with spatial data. Here, we evaluated the performance of Random Forest in upscaling airborne LiDAR (Light Detection and Ranging)-based carbon estimates compared to the stratification approach over a 16-million hectare focal area of the Western Amazon. We considered two runs of Random Forest, both with and without spatial contextual modeling by including—in the latter case—x, and y position directly in the model. In each case, we set aside 8 million hectares (i.e., half of the focal area) for validation; this rigorous test of Random Forest went above and beyond the internal validation normally compiled by the algorithm (i.e., called “out-of-bag”), which proved insufficient for this spatial application. In this heterogeneous region of Northern Peru, the model with spatial context was the best preforming run of Random Forest, and explained 59% of LiDAR-based carbon estimates within the validation area, compared to 37% for stratification or 43% by Random Forest without spatial context. With the 60% improvement in explained variation, RMSE against validation LiDAR samples improved from 33 to 26 Mg C ha−1 when using Random Forest with spatial context. Our results suggest that spatial context should be considered when using Random Forest, and that doing so may result in substantially improved carbon stock modeling for purposes of climate change mitigation. PMID:24489686

Forest resources of the Umatilla National Forest.

Treesearch

Glenn A. Christensen; Paul Dunham; David C. Powell; Bruce. Hiserote

2007-01-01

Current resource statistics for the Umatilla National Forest, based on two separate inventories conducted in 1993–96 and in 1997–2002, are presented in this report. Currently on the Umatilla National Forest, 89 percent of the land area is classified as forest land. The predominant forest type is grand fir (26 percent of forested acres) followed by the interior Douglas-...

Automatic quality control in clinical (1)H MRSI of brain cancer.

PubMed

Pedrosa de Barros, Nuno; McKinley, Richard; Knecht, Urspeter; Wiest, Roland; Slotboom, Johannes

2016-05-01

MRSI grids frequently show spectra with poor quality, mainly because of the high sensitivity of MRS to field inhomogeneities. These poor quality spectra are prone to quantification and/or interpretation errors that can have a significant impact on the clinical use of spectroscopic data. Therefore, quality control of the spectra should always precede their clinical use. When performed manually, quality assessment of MRSI spectra is not only a tedious and time-consuming task, but is also affected by human subjectivity. Consequently, automatic, fast and reliable methods for spectral quality assessment are of utmost interest. In this article, we present a new random forest-based method for automatic quality assessment of (1)H MRSI brain spectra, which uses a new set of MRS signal features. The random forest classifier was trained on spectra from 40 MRSI grids that were classified as acceptable or non-acceptable by two expert spectroscopists. To account for the effects of intra-rater reliability, each spectrum was rated for quality three times by each rater. The automatic method classified these spectra with an area under the curve (AUC) of 0.976. Furthermore, in the subset of spectra containing only the cases that were classified every time in the same way by the spectroscopists, an AUC of 0.998 was obtained. Feature importance for the classification was also evaluated. Frequency domain skewness and kurtosis, as well as time domain signal-to-noise ratios (SNRs) in the ranges 50-75 ms and 75-100 ms, were the most important features. Given that the method is able to assess a whole MRSI grid faster than a spectroscopist (approximately 3 s versus approximately 3 min), and without loss of accuracy (agreement between classifier trained with just one session and any of the other labelling sessions, 89.88%; agreement between any two labelling sessions, 89.03%), the authors suggest its implementation in the clinical routine. The method presented in this article was implemented in jMRUI's SpectrIm plugin. Copyright © 2016 John Wiley & Sons, Ltd.

Reading Profiles in Multi-Site Data With Missingness.

PubMed

Eckert, Mark A; Vaden, Kenneth I; Gebregziabher, Mulugeta

2018-01-01

Children with reading disability exhibit varied deficits in reading and cognitive abilities that contribute to their reading comprehension problems. Some children exhibit primary deficits in phonological processing, while others can exhibit deficits in oral language and executive functions that affect comprehension. This behavioral heterogeneity is problematic when missing data prevent the characterization of different reading profiles, which often occurs in retrospective data sharing initiatives without coordinated data collection. Here we show that reading profiles can be reliably identified based on Random Forest classification of incomplete behavioral datasets, after the missForest method is used to multiply impute missing values. Results from simulation analyses showed that reading profiles could be accurately classified across degrees of missingness (e.g., ∼5% classification error for 30% missingness across the sample). The application of missForest to a real multi-site dataset with missingness ( n = 924) showed that reading disability profiles significantly and consistently differed in reading and cognitive abilities for cases with and without missing data. The results of validation analyses indicated that the reading profiles (cases with and without missing data) exhibited significant differences for an independent set of behavioral variables that were not used to classify reading profiles. Together, the results show how multiple imputation can be applied to the classification of cases with missing data and can increase the integrity of results from multi-site open access datasets.

Local receptive field constrained stacked sparse autoencoder for classification of hyperspectral images.

PubMed

Wan, Xiaoqing; Zhao, Chunhui

2017-06-01

As a competitive machine learning algorithm, the stacked sparse autoencoder (SSA) has achieved outstanding popularity in exploiting high-level features for classification of hyperspectral images (HSIs). In general, in the SSA architecture, the nodes between adjacent layers are fully connected and need to be iteratively fine-tuned during the pretraining stage; however, the nodes of previous layers further away may be less likely to have a dense correlation to the given node of subsequent layers. Therefore, to reduce the classification error and increase the learning rate, this paper proposes the general framework of locally connected SSA; that is, the biologically inspired local receptive field (LRF) constrained SSA architecture is employed to simultaneously characterize the local correlations of spectral features and extract high-level feature representations of hyperspectral data. In addition, the appropriate receptive field constraint is concurrently updated by measuring the spatial distances from the neighbor nodes to the corresponding node. Finally, the efficient random forest classifier is cascaded to the last hidden layer of the SSA architecture as a benchmark classifier. Experimental results on two real HSI datasets demonstrate that the proposed hierarchical LRF constrained stacked sparse autoencoder and random forest (SSARF) provides encouraging results with respect to other contrastive methods, for instance, the improvements of overall accuracy in a range of 0.72%-10.87% for the Indian Pines dataset and 0.74%-7.90% for the Kennedy Space Center dataset; moreover, it generates lower running time compared with the result provided by similar SSARF based methodology.

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

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 %.

Method: automatic segmentation of mitochondria utilizing patch classification, contour pair classification, and automatically seeded level sets

PubMed Central

2012-01-01

Background While progress has been made to develop automatic segmentation techniques for mitochondria, there remains a need for more accurate and robust techniques to delineate mitochondria in serial blockface scanning electron microscopic data. Previously developed texture based methods are limited for solving this problem because texture alone is often not sufficient to identify mitochondria. This paper presents a new three-step method, the Cytoseg process, for automated segmentation of mitochondria contained in 3D electron microscopic volumes generated through serial block face scanning electron microscopic imaging. The method consists of three steps. The first is a random forest patch classification step operating directly on 2D image patches. The second step consists of contour-pair classification. At the final step, we introduce a method to automatically seed a level set operation with output from previous steps. Results We report accuracy of the Cytoseg process on three types of tissue and compare it to a previous method based on Radon-Like Features. At step 1, we show that the patch classifier identifies mitochondria texture but creates many false positive pixels. At step 2, our contour processing step produces contours and then filters them with a second classification step, helping to improve overall accuracy. We show that our final level set operation, which is automatically seeded with output from previous steps, helps to smooth the results. Overall, our results show that use of contour pair classification and level set operations improve segmentation accuracy beyond patch classification alone. We show that the Cytoseg process performs well compared to another modern technique based on Radon-Like Features. Conclusions We demonstrated that texture based methods for mitochondria segmentation can be enhanced with multiple steps that form an image processing pipeline. While we used a random-forest based patch classifier to recognize texture, it would be possible to replace this with other texture identifiers, and we plan to explore this in future work. PMID:22321695

Heterogeneous classifier fusion for ligand-based virtual screening: or, how decision making by committee can be a good thing.

PubMed

Riniker, Sereina; Fechner, Nikolas; Landrum, Gregory A

2013-11-25

The concept of data fusion - the combination of information from different sources describing the same object with the expectation to generate a more accurate representation - has found application in a very broad range of disciplines. In the context of ligand-based virtual screening (VS), data fusion has been applied to combine knowledge from either different active molecules or different fingerprints to improve similarity search performance. Machine-learning (ML) methods based on fusion of multiple homogeneous classifiers, in particular random forests, have also been widely applied in the ML literature. The heterogeneous version of classifier fusion - fusing the predictions from different model types - has been less explored. Here, we investigate heterogeneous classifier fusion for ligand-based VS using three different ML methods, RF, naïve Bayes (NB), and logistic regression (LR), with four 2D fingerprints, atom pairs, topological torsions, RDKit fingerprint, and circular fingerprint. The methods are compared using a previously developed benchmarking platform for 2D fingerprints which is extended to ML methods in this article. The original data sets are filtered for difficulty, and a new set of challenging data sets from ChEMBL is added. Data sets were also generated for a second use case: starting from a small set of related actives instead of diverse actives. The final fused model consistently outperforms the other approaches across the broad variety of targets studied, indicating that heterogeneous classifier fusion is a very promising approach for ligand-based VS. The new data sets together with the adapted source code for ML methods are provided in the Supporting Information .

FRAGMENTATION OF CONTINENTAL UNITES STATES FORESTS

EPA Science Inventory

We report a multiple-scale analysis of forest fragmentation based on 30-m land-cover maps for the conterminous United States. Each 0.09-ha unit of forest was classified according to fragmentation indices measured within the surrounding landscape, for five landscape sizes from 2....

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

Automated segmentation of thyroid gland on CT images with multi-atlas label fusion and random classification forest

NASA Astrophysics Data System (ADS)

Liu, Jiamin; Chang, Kevin; Kim, Lauren; Turkbey, Evrim; Lu, Le; Yao, Jianhua; Summers, Ronald

2015-03-01

The thyroid gland plays an important role in clinical practice, especially for radiation therapy treatment planning. For patients with head and neck cancer, radiation therapy requires a precise delineation of the thyroid gland to be spared on the pre-treatment planning CT images to avoid thyroid dysfunction. In the current clinical workflow, the thyroid gland is normally manually delineated by radiologists or radiation oncologists, which is time consuming and error prone. Therefore, a system for automated segmentation of the thyroid is desirable. However, automated segmentation of the thyroid is challenging because the thyroid is inhomogeneous and surrounded by structures that have similar intensities. In this work, the thyroid gland segmentation is initially estimated by multi-atlas label fusion algorithm. The segmentation is refined by supervised statistical learning based voxel labeling with a random forest algorithm. Multiatlas label fusion (MALF) transfers expert-labeled thyroids from atlases to a target image using deformable registration. Errors produced by label transfer are reduced by label fusion that combines the results produced by all atlases into a consensus solution. Then, random forest (RF) employs an ensemble of decision trees that are trained on labeled thyroids to recognize features. The trained forest classifier is then applied to the thyroid estimated from the MALF by voxel scanning to assign the class-conditional probability. Voxels from the expert-labeled thyroids in CT volumes are treated as positive classes; background non-thyroid voxels as negatives. We applied this automated thyroid segmentation system to CT scans of 20 patients. The results showed that the MALF achieved an overall 0.75 Dice Similarity Coefficient (DSC) and the RF classification further improved the DSC to 0.81.

Forested land cover classification on the Cumberland Plateau, Jackson County, Alabama: a comparison of Landsat ETM+ and SPOT5 images

Treesearch

Yong Wang; Shanta Parajuli; Callie Schweitzer; Glendon Smalley; Dawn Lemke; Wubishet Tadesse; Xiongwen Chen

2010-01-01

Forest cover classifications focus on the overall growth form (physiognomy) of the community, dominant vegetation, and species composition of the existing forest. Accurately classifying the forest cover type is important for forest inventory and silviculture. We compared classification accuracy based on Landsat Enhanced Thematic Mapper Plus (Landsat ETM+) and Satellite...

The ASAS-SN catalogue of variable stars I: The Serendipitous Survey

NASA Astrophysics Data System (ADS)

Jayasinghe, T.; Kochanek, C. S.; Stanek, K. Z.; Shappee, B. J.; Holoien, T. W.-S.; Thompson, Toda A.; Prieto, J. L.; Dong, Subo; Pawlak, M.; Shields, J. V.; Pojmanski, G.; Otero, S.; Britt, C. A.; Will, D.

2018-07-01

The All-Sky Automated Survey for Supernovae (ASAS-SN) is the first optical survey to routinely monitor the whole sky with a cadence of ˜2-3 d down to V ≲ 17 mag. ASAS-SN has monitored the whole sky since 2014, collecting ˜100-500 epochs of observations per field. The V-band light curves for candidate variables identified during the search for supernovae are classified using a random forest classifier and visually verified. We present a catalogue of 66 179 bright, new variable stars discovered during our search for supernovae, including 27 479 periodic variables and 38 700 irregular variables. V-band light curves for the ASAS-SN variables are available through the ASAS-SN variable stars data base (https://asas-sn.osu.edu/variables). The data base will begin to include the light curves of known variable stars in the near future along with the results for a systematic, all-sky variability survey.

Classifying clinical notes with pain assessment using machine learning.

PubMed

Fodeh, Samah Jamal; Finch, Dezon; Bouayad, Lina; Luther, Stephen L; Ling, Han; Kerns, Robert D; Brandt, Cynthia

2017-12-26

Pain is a significant public health problem, affecting millions of people in the USA. Evidence has highlighted that patients with chronic pain often suffer from deficits in pain care quality (PCQ) including pain assessment, treatment, and reassessment. Currently, there is no intelligent and reliable approach to identify PCQ indicators inelectronic health records (EHR). Hereby, we used unstructured text narratives in the EHR to derive pain assessment in clinical notes for patients with chronic pain. Our dataset includes patients with documented pain intensity rating ratings > = 4 and initial musculoskeletal diagnoses (MSD) captured by (ICD-9-CM codes) in fiscal year 2011 and a minimal 1 year of follow-up (follow-up period is 3-yr maximum); with complete data on key demographic variables. A total of 92 patients with 1058 notes was used. First, we manually annotated qualifiers and descriptors of pain assessment using the annotation schema that we previously developed. Second, we developed a reliable classifier for indicators of pain assessment in clinical note. Based on our annotation schema, we found variations in documenting the subclasses of pain assessment. In positive notes, providers mostly documented assessment of pain site (67%) and intensity of pain (57%), followed by persistence (32%). In only 27% of positive notes, did providers document a presumed etiology for the pain complaint or diagnosis. Documentation of patients' reports of factors that aggravate pain was only present in 11% of positive notes. Random forest classifier achieved the best performance labeling clinical notes with pain assessment information, compared to other classifiers; 94, 95, 94, and 94% was observed in terms of accuracy, PPV, F1-score, and AUC, respectively. Despite the wide spectrum of research that utilizes machine learning in many clinical applications, none explored using these methods for pain assessment research. In addition, previous studies using large datasets to detect and analyze characteristics of patients with various types of pain have relied exclusively on billing and coded data as the main source of information. This study, in contrast, harnessed unstructured narrative text data from the EHR to detect pain assessment clinical notes. We developed a Random forest classifier to identify clinical notes with pain assessment information. Compared to other classifiers, ours achieved the best results in most of the reported metrics. Graphical abstract Framework for detecting pain assessment in clinical notes.

Calibrating random forests for probability estimation.

PubMed

Dankowski, Theresa; Ziegler, Andreas

2016-09-30

Probabilities can be consistently estimated using random forests. It is, however, unclear how random forests should be updated to make predictions for other centers or at different time points. In this work, we present two approaches for updating random forests for probability estimation. The first method has been proposed by Elkan and may be used for updating any machine learning approach yielding consistent probabilities, so-called probability machines. The second approach is a new strategy specifically developed for random forests. Using the terminal nodes, which represent conditional probabilities, the random forest is first translated to logistic regression models. These are, in turn, used for re-calibration. The two updating strategies were compared in a simulation study and are illustrated with data from the German Stroke Study Collaboration. In most simulation scenarios, both methods led to similar improvements. In the simulation scenario in which the stricter assumptions of Elkan's method were not met, the logistic regression-based re-calibration approach for random forests outperformed Elkan's method. It also performed better on the stroke data than Elkan's method. The strength of Elkan's method is its general applicability to any probability machine. However, if the strict assumptions underlying this approach are not met, the logistic regression-based approach is preferable for updating random forests for probability estimation. © 2016 The Authors. Statistics in Medicine Published by John Wiley & Sons Ltd. © 2016 The Authors. Statistics in Medicine Published by John Wiley & Sons Ltd.

Discriminating Canopy Structural Types from Optical Properties using AVIRIS Data in the Sierra National Forest in Central California

NASA Astrophysics Data System (ADS)

Huesca Martinez, M.; Garcia, M.; Roth, K. L.; Casas, A.; Ustin, S.

2015-12-01

There is a well-established need within the remote sensing community for improved estimation of canopy structure and understanding of its influence on the retrieval of leaf biochemical properties. The aim of this project was to evaluate the estimation of structural properties directly from hyperspectral data, with the broader goal that these might be used to constrain retrievals of canopy chemistry. We used NASA's Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) to discriminate different canopy structural types, defined in terms of biomass, canopy height and vegetation complexity, and compared them to estimates of these properties measured by LiDAR data. We tested a large number of optical metrics, including single narrow band reflectance and 1st derivative, sub-pixel cover fractions, narrow-band indices, spectral absorption features, and Principal Component Analysis components. Canopy structural types were identified and classified from different forest types by integrating structural traits measured by optical metrics using the Random Forest (RF) classifier. The classification accuracy was above 70% in most of the vegetation scenarios. The best overall accuracy was achieved for hardwood forest (>80% accuracy) and the lowest accuracy was found in mixed forest (~70% accuracy). Furthermore, similarly high accuracy was found when the RF classifier was applied to a spatially independent dataset, showing significant portability for the method used. Results show that all spectral regions played a role in canopy structure assessment, thus the whole spectrum is required. Furthermore, optical metrics derived from AVIRIS proved to be a powerful technique for structural attribute mapping. This research illustrates the potential for using optical properties to distinguish several canopy structural types in different forest types, and these may be used to constrain quantitative measurements of absorbing properties in future research.

Pseudo CT estimation from MRI using patch-based random forest

NASA Astrophysics Data System (ADS)

Yang, Xiaofeng; Lei, Yang; Shu, Hui-Kuo; Rossi, Peter; Mao, Hui; Shim, Hyunsuk; Curran, Walter J.; Liu, Tian

2017-02-01

Recently, MR simulators gain popularity because of unnecessary radiation exposure of CT simulators being used in radiation therapy planning. We propose a method for pseudo CT estimation from MR images based on a patch-based random forest. Patient-specific anatomical features are extracted from the aligned training images and adopted as signatures for each voxel. The most robust and informative features are identified using feature selection to train the random forest. The well-trained random forest is used to predict the pseudo CT of a new patient. This prediction technique was tested with human brain images and the prediction accuracy was assessed using the original CT images. Peak signal-to-noise ratio (PSNR) and feature similarity (FSIM) indexes were used to quantify the differences between the pseudo and original CT images. The experimental results showed the proposed method could accurately generate pseudo CT images from MR images. In summary, we have developed a new pseudo CT prediction method based on patch-based random forest, demonstrated its clinical feasibility, and validated its prediction accuracy. This pseudo CT prediction technique could be a useful tool for MRI-based radiation treatment planning and attenuation correction in a PET/MRI scanner.

Machine- z: Rapid machine-learned redshift indicator for Swift gamma-ray bursts

DOE PAGES

Ukwatta, T. N.; Wozniak, P. R.; Gehrels, N.

2016-03-08

Studies of high-redshift gamma-ray bursts (GRBs) provide important information about the early Universe such as the rates of stellar collapsars and mergers, the metallicity content, constraints on the re-ionization period, and probes of the Hubble expansion. Rapid selection of high-z candidates from GRB samples reported in real time by dedicated space missions such as Swift is the key to identifying the most distant bursts before the optical afterglow becomes too dim to warrant a good spectrum. Here, we introduce ‘machine-z’, a redshift prediction algorithm and a ‘high-z’ classifier for Swift GRBs based on machine learning. Our method relies exclusively onmore » canonical data commonly available within the first few hours after the GRB trigger. Using a sample of 284 bursts with measured redshifts, we trained a randomized ensemble of decision trees (random forest) to perform both regression and classification. Cross-validated performance studies show that the correlation coefficient between machine-z predictions and the true redshift is nearly 0.6. At the same time, our high-z classifier can achieve 80 per cent recall of true high-redshift bursts, while incurring a false positive rate of 20 per cent. With 40 per cent false positive rate the classifier can achieve ~100 per cent recall. As a result, the most reliable selection of high-redshift GRBs is obtained by combining predictions from both the high-z classifier and the machine-z regressor.« less

Automated Scoring of L2 Spoken English with Random Forests

ERIC Educational Resources Information Center

Kobayashi, Yuichiro; Abe, Mariko

2016-01-01

The purpose of the present study is to assess second language (L2) spoken English using automated scoring techniques. Automated scoring aims to classify a large set of learners' oral performance data into a small number of discrete oral proficiency levels. In automated scoring, objectively measurable features such as the frequencies of lexical and…

voomDDA: discovery of diagnostic biomarkers and classification of RNA-seq data.

PubMed

Zararsiz, Gokmen; Goksuluk, Dincer; Klaus, Bernd; Korkmaz, Selcuk; Eldem, Vahap; Karabulut, Erdem; Ozturk, Ahmet

2017-01-01

RNA-Seq is a recent and efficient technique that uses the capabilities of next-generation sequencing technology for characterizing and quantifying transcriptomes. One important task using gene-expression data is to identify a small subset of genes that can be used to build diagnostic classifiers particularly for cancer diseases. Microarray based classifiers are not directly applicable to RNA-Seq data due to its discrete nature. Overdispersion is another problem that requires careful modeling of mean and variance relationship of the RNA-Seq data. In this study, we present voomDDA classifiers: variance modeling at the observational level (voom) extensions of the nearest shrunken centroids (NSC) and the diagonal discriminant classifiers. VoomNSC is one of these classifiers and brings voom and NSC approaches together for the purpose of gene-expression based classification. For this purpose, we propose weighted statistics and put these weighted statistics into the NSC algorithm. The VoomNSC is a sparse classifier that models the mean-variance relationship using the voom method and incorporates voom's precision weights into the NSC classifier via weighted statistics. A comprehensive simulation study was designed and four real datasets are used for performance assessment. The overall results indicate that voomNSC performs as the sparsest classifier. It also provides the most accurate results together with power-transformed Poisson linear discriminant analysis, rlog transformed support vector machines and random forests algorithms. In addition to prediction purposes, the voomNSC classifier can be used to identify the potential diagnostic biomarkers for a condition of interest. Through this work, statistical learning methods proposed for microarrays can be reused for RNA-Seq data. An interactive web application is freely available at http://www.biosoft.hacettepe.edu.tr/voomDDA/.

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.

Floodplain Mapping for the Continental United States Using Machine Learning Techniques and Watershed Characteristics

NASA Astrophysics Data System (ADS)

Jafarzadegan, K.; Merwade, V.; Saksena, S.

2017-12-01

Using conventional hydrodynamic methods for floodplain mapping in large-scale and data-scarce regions is problematic due to the high cost of these methods, lack of reliable data and uncertainty propagation. In this study a new framework is proposed to generate 100-year floodplains for any gauged or ungauged watershed across the United States (U.S.). This framework uses Flood Insurance Rate Maps (FIRMs), topographic, climatic and land use data which are freely available for entire U.S. for floodplain mapping. The framework consists of three components, including a Random Forest classifier for watershed classification, a Probabilistic Threshold Binary Classifier (PTBC) for generating the floodplains, and a lookup table for linking the Random Forest classifier to the PTBC. The effectiveness and reliability of the proposed framework is tested on 145 watersheds from various geographical locations in the U.S. The validation results show that around 80 percent of total watersheds are predicted well, 14 percent have acceptable fit and less than five percent are predicted poorly compared to FIRMs. Another advantage of this framework is its ability in generating floodplains for all small rivers and tributaries. Due to the high accuracy and efficiency of this framework, it can be used as a preliminary decision making tool to generate 100-year floodplain maps for data-scarce regions and all tributaries where hydrodynamic methods are difficult to use.

Automatic classification of time-variable X-ray sources

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

Lo, Kitty K.; Farrell, Sean; Murphy, Tara

2014-05-01

To maximize the discovery potential of future synoptic surveys, especially in the field of transient science, it will be necessary to use automatic classification to identify some of the astronomical sources. The data mining technique of supervised classification is suitable for this problem. Here, we present a supervised learning method to automatically classify variable X-ray sources in the Second XMM-Newton Serendipitous Source Catalog (2XMMi-DR2). Random Forest is our classifier of choice since it is one of the most accurate learning algorithms available. Our training set consists of 873 variable sources and their features are derived from time series, spectra, andmore » other multi-wavelength contextual information. The 10 fold cross validation accuracy of the training data is ∼97% on a 7 class data set. We applied the trained classification model to 411 unknown variable 2XMM sources to produce a probabilistically classified catalog. Using the classification margin and the Random Forest derived outlier measure, we identified 12 anomalous sources, of which 2XMM J180658.7–500250 appears to be the most unusual source in the sample. Its X-ray spectra is suggestive of a ultraluminous X-ray source but its variability makes it highly unusual. Machine-learned classification and anomaly detection will facilitate scientific discoveries in the era of all-sky surveys.« less

Peculiarities of use of ECOC and AdaBoost based classifiers for thematic processing of hyperspectral data

NASA Astrophysics Data System (ADS)

Dementev, A. O.; Dmitriev, E. V.; Kozoderov, V. V.; Egorov, V. D.

2017-10-01

Hyperspectral imaging is up-to-date promising technology widely applied for the accurate thematic mapping. The presence of a large number of narrow survey channels allows us to use subtle differences in spectral characteristics of objects and to make a more detailed classification than in the case of using standard multispectral data. The difficulties encountered in the processing of hyperspectral images are usually associated with the redundancy of spectral information which leads to the problem of the curse of dimensionality. Methods currently used for recognizing objects on multispectral and hyperspectral images are usually based on standard base supervised classification algorithms of various complexity. Accuracy of these algorithms can be significantly different depending on considered classification tasks. In this paper we study the performance of ensemble classification methods for the problem of classification of the forest vegetation. Error correcting output codes and boosting are tested on artificial data and real hyperspectral images. It is demonstrates, that boosting gives more significant improvement when used with simple base classifiers. The accuracy in this case in comparable the error correcting output code (ECOC) classifier with Gaussian kernel SVM base algorithm. However the necessity of boosting ECOC with Gaussian kernel SVM is questionable. It is demonstrated, that selected ensemble classifiers allow us to recognize forest species with high enough accuracy which can be compared with ground-based forest inventory data.

A comparison of the conditional inference survival forest model to random survival forests based on a simulation study as well as on two applications with time-to-event data.

PubMed

Nasejje, Justine B; Mwambi, Henry; Dheda, Keertan; Lesosky, Maia

2017-07-28

Random survival forest (RSF) models have been identified as alternative methods to the Cox proportional hazards model in analysing time-to-event data. These methods, however, have been criticised for the bias that results from favouring covariates with many split-points and hence conditional inference forests for time-to-event data have been suggested. Conditional inference forests (CIF) are known to correct the bias in RSF models by separating the procedure for the best covariate to split on from that of the best split point search for the selected covariate. In this study, we compare the random survival forest model to the conditional inference model (CIF) using twenty-two simulated time-to-event datasets. We also analysed two real time-to-event datasets. The first dataset is based on the survival of children under-five years of age in Uganda and it consists of categorical covariates with most of them having more than two levels (many split-points). The second dataset is based on the survival of patients with extremely drug resistant tuberculosis (XDR TB) which consists of mainly categorical covariates with two levels (few split-points). The study findings indicate that the conditional inference forest model is superior to random survival forest models in analysing time-to-event data that consists of covariates with many split-points based on the values of the bootstrap cross-validated estimates for integrated Brier scores. However, conditional inference forests perform comparably similar to random survival forests models in analysing time-to-event data consisting of covariates with fewer split-points. Although survival forests are promising methods in analysing time-to-event data, it is important to identify the best forest model for analysis based on the nature of covariates of the dataset in question.

Mapping the distribution of the main host for plague in a complex landscape in Kazakhstan: An object-based approach using SPOT-5 XS, Landsat 7 ETM+, SRTM and multiple Random Forests

NASA Astrophysics Data System (ADS)

Wilschut, L. I.; Addink, E. A.; Heesterbeek, J. A. P.; Dubyanskiy, V. M.; Davis, S. A.; Laudisoit, A.; Begon, M.; Burdelov, L. A.; Atshabar, B. B.; de Jong, S. M.

2013-08-01

Plague is a zoonotic infectious disease present in great gerbil populations in Kazakhstan. Infectious disease dynamics are influenced by the spatial distribution of the carriers (hosts) of the disease. The great gerbil, the main host in our study area, lives in burrows, which can be recognized on high resolution satellite imagery. In this study, using earth observation data at various spatial scales, we map the spatial distribution of burrows in a semi-desert landscape. The study area consists of various landscape types. To evaluate whether identification of burrows by classification is possible in these landscape types, the study area was subdivided into eight landscape units, on the basis of Landsat 7 ETM+ derived Tasselled Cap Greenness and Brightness, and SRTM derived standard deviation in elevation. In the field, 904 burrows were mapped. Using two segmented 2.5 m resolution SPOT-5 XS satellite scenes, reference object sets were created. Random Forests were built for both SPOT scenes and used to classify the images. Additionally, a stratified classification was carried out, by building separate Random Forests per landscape unit. Burrows were successfully classified in all landscape units. In the ‘steppe on floodplain’ areas, classification worked best: producer's and user's accuracy in those areas reached 88% and 100%, respectively. In the ‘floodplain’ areas with a more heterogeneous vegetation cover, classification worked least well; there, accuracies were 86 and 58% respectively. Stratified classification improved the results in all landscape units where comparison was possible (four), increasing kappa coefficients by 13, 10, 9 and 1%, respectively. In this study, an innovative stratification method using high- and medium resolution imagery was applied in order to map host distribution on a large spatial scale. The burrow maps we developed will help to detect changes in the distribution of great gerbil populations and, moreover, serve as a unique empirical data set which can be used as input for epidemiological plague models. This is an important step in understanding the dynamics of plague.

Fuzzy association rule mining and classification for the prediction of malaria in South Korea.

PubMed

Buczak, Anna L; Baugher, Benjamin; Guven, Erhan; Ramac-Thomas, Liane C; Elbert, Yevgeniy; Babin, Steven M; Lewis, Sheri H

2015-06-18

Malaria is the world's most prevalent vector-borne disease. Accurate prediction of malaria outbreaks may lead to public health interventions that mitigate disease morbidity and mortality. We describe an application of a method for creating prediction models utilizing Fuzzy Association Rule Mining to extract relationships between epidemiological, meteorological, climatic, and socio-economic data from Korea. These relationships are in the form of rules, from which the best set of rules is automatically chosen and forms a classifier. Two classifiers have been built and their results fused to become a malaria prediction model. Future malaria cases are predicted as Low, Medium or High, where these classes are defined as a total of 0-2, 3-16, and above 17 cases, respectively, for a region in South Korea during a two-week period. Based on user recommendations, HIGH is considered an outbreak. Model accuracy is described by Positive Predictive Value (PPV), Sensitivity, and F-score for each class, computed on test data not previously used to develop the model. For predictions made 7-8 weeks in advance, model PPV and Sensitivity are 0.842 and 0.681, respectively, for the HIGH classes. The F0.5 and F3 scores (which combine PPV and Sensitivity) are 0.804 and 0.694, respectively, for the HIGH classes. The overall FARM results (as measured by F-scores) are significantly better than those obtained by Decision Tree, Random Forest, Support Vector Machine, and Holt-Winters methods for the HIGH class. For the Medium class, Random Forest and FARM obtain comparable results, with FARM being better at F0.5, and Random Forest obtaining a higher F3. A previously described method for creating disease prediction models has been modified and extended to build models for predicting malaria. In addition, some new input variables were used, including indicators of intervention measures. The South Korea malaria prediction models predict Low, Medium or High cases 7-8 weeks in the future. This paper demonstrates that our data driven approach can be used for the prediction of different diseases.

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.

Classifier ensemble construction with rotation forest to improve medical diagnosis performance of machine learning algorithms.

PubMed

Ozcift, Akin; Gulten, Arif

2011-12-01

Improving accuracies of machine learning algorithms is vital in designing high performance computer-aided diagnosis (CADx) systems. Researches have shown that a base classifier performance might be enhanced by ensemble classification strategies. In this study, we construct rotation forest (RF) ensemble classifiers of 30 machine learning algorithms to evaluate their classification performances using Parkinson's, diabetes and heart diseases from literature. While making experiments, first the feature dimension of three datasets is reduced using correlation based feature selection (CFS) algorithm. Second, classification performances of 30 machine learning algorithms are calculated for three datasets. Third, 30 classifier ensembles are constructed based on RF algorithm to assess performances of respective classifiers with the same disease data. All the experiments are carried out with leave-one-out validation strategy and the performances of the 60 algorithms are evaluated using three metrics; classification accuracy (ACC), kappa error (KE) and area under the receiver operating characteristic (ROC) curve (AUC). Base classifiers succeeded 72.15%, 77.52% and 84.43% average accuracies for diabetes, heart and Parkinson's datasets, respectively. As for RF classifier ensembles, they produced average accuracies of 74.47%, 80.49% and 87.13% for respective diseases. RF, a newly proposed classifier ensemble algorithm, might be used to improve accuracy of miscellaneous machine learning algorithms to design advanced CADx systems. Copyright © 2011 Elsevier Ireland Ltd. All rights reserved.

Identification of an Efficient Gene Expression Panel for Glioblastoma Classification

PubMed Central

Zelaya, Ivette; Laks, Dan R.; Zhao, Yining; Kawaguchi, Riki; Gao, Fuying; Kornblum, Harley I.; Coppola, Giovanni

2016-01-01

We present here a novel genetic algorithm-based random forest (GARF) modeling technique that enables a reduction in the complexity of large gene disease signatures to highly accurate, greatly simplified gene panels. When applied to 803 glioblastoma multiforme samples, this method allowed the 840-gene Verhaak et al. gene panel (the standard in the field) to be reduced to a 48-gene classifier, while retaining 90.91% classification accuracy, and outperforming the best available alternative methods. Additionally, using this approach we produced a 32-gene panel which allows for better consistency between RNA-seq and microarray-based classifications, improving cross-platform classification retention from 69.67% to 86.07%. A webpage producing these classifications is available at http://simplegbm.semel.ucla.edu. PMID:27855170

Proteomics Versus Clinical Data and Stochastic Local Search Based Feature Selection for Acute Myeloid Leukemia Patients' Classification.

PubMed

Chebouba, Lokmane; Boughaci, Dalila; Guziolowski, Carito

2018-06-04

The use of data issued from high throughput technologies in drug target problems is widely widespread during the last decades. This study proposes a meta-heuristic framework using stochastic local search (SLS) combined with random forest (RF) where the aim is to specify the most important genes and proteins leading to the best classification of Acute Myeloid Leukemia (AML) patients. First we use a stochastic local search meta-heuristic as a feature selection technique to select the most significant proteins to be used in the classification task step. Then we apply RF to classify new patients into their corresponding classes. The evaluation technique is to run the RF classifier on the training data to get a model. Then, we apply this model on the test data to find the appropriate class. We use as metrics the balanced accuracy (BAC) and the area under the receiver operating characteristic curve (AUROC) to measure the performance of our model. The proposed method is evaluated on the dataset issued from DREAM 9 challenge. The comparison is done with a pure random forest (without feature selection), and with the two best ranked results of the DREAM 9 challenge. We used three types of data: only clinical data, only proteomics data, and finally clinical and proteomics data combined. The numerical results show that the highest scores are obtained when using clinical data alone, and the lowest is obtained when using proteomics data alone. Further, our method succeeds in finding promising results compared to the methods presented in the DREAM challenge.

Multivariate classification with random forests for gravitational wave searches of black hole binary coalescence

NASA Astrophysics Data System (ADS)

Baker, Paul T.; Caudill, Sarah; Hodge, Kari A.; Talukder, Dipongkar; Capano, Collin; Cornish, Neil J.

2015-03-01

Searches for gravitational waves produced by coalescing black hole binaries with total masses ≳25 M⊙ use matched filtering with templates of short duration. Non-Gaussian noise bursts in gravitational wave detector data can mimic short signals and limit the sensitivity of these searches. Previous searches have relied on empirically designed statistics incorporating signal-to-noise ratio and signal-based vetoes to separate gravitational wave candidates from noise candidates. We report on sensitivity improvements achieved using a multivariate candidate ranking statistic derived from a supervised machine learning algorithm. We apply the random forest of bagged decision trees technique to two separate searches in the high mass (≳25 M⊙ ) parameter space. For a search which is sensitive to gravitational waves from the inspiral, merger, and ringdown of binary black holes with total mass between 25 M⊙ and 100 M⊙ , we find sensitive volume improvements as high as 70±13%-109±11% when compared to the previously used ranking statistic. For a ringdown-only search which is sensitive to gravitational waves from the resultant perturbed intermediate mass black hole with mass roughly between 10 M⊙ and 600 M⊙ , we find sensitive volume improvements as high as 61±4%-241±12% when compared to the previously used ranking statistic. We also report how sensitivity improvements can differ depending on mass regime, mass ratio, and available data quality information. Finally, we describe the techniques used to tune and train the random forest classifier that can be generalized to its use in other searches for gravitational waves.

Classifying plant series-level forest potential types: methods for subbasins sampled in the midscale assessment of the interior Columbia basin.

Treesearch

Paul F. Hessburg; Bradley G. Smith; Scott D. Kreiter; Craig A. Miller; Cecilia H. McNicoll; Michele. Wasienko-Holland

2000-01-01

In the interior Columbia River basin midscale ecological assessment, we mapped and characterized historical and current vegetation composition and structure of 337 randomly sampled subwatersheds (9500 ha average size) in 43 subbasins (404 000 ha average size). We compared landscape patterns, vegetation structure and composition, and landscape vulnerability to wildfires...

Retinal layer segmentation of macular OCT images using boundary classification

PubMed Central

Lang, Andrew; Carass, Aaron; Hauser, Matthew; Sotirchos, Elias S.; Calabresi, Peter A.; Ying, Howard S.; Prince, Jerry L.

2013-01-01

Optical coherence tomography (OCT) has proven to be an essential imaging modality for ophthalmology and is proving to be very important in neurology. OCT enables high resolution imaging of the retina, both at the optic nerve head and the macula. Macular retinal layer thicknesses provide useful diagnostic information and have been shown to correlate well with measures of disease severity in several diseases. Since manual segmentation of these layers is time consuming and prone to bias, automatic segmentation methods are critical for full utilization of this technology. In this work, we build a random forest classifier to segment eight retinal layers in macular cube images acquired by OCT. The random forest classifier learns the boundary pixels between layers, producing an accurate probability map for each boundary, which is then processed to finalize the boundaries. Using this algorithm, we can accurately segment the entire retina contained in the macular cube to an accuracy of at least 4.3 microns for any of the nine boundaries. Experiments were carried out on both healthy and multiple sclerosis subjects, with no difference in the accuracy of our algorithm found between the groups. PMID:23847738

Vertebral degenerative disc disease severity evaluation using random forest classification

NASA Astrophysics Data System (ADS)

Munoz, Hector E.; Yao, Jianhua; Burns, Joseph E.; Pham, Yasuyuki; Stieger, James; Summers, Ronald M.

2014-03-01

Degenerative disc disease (DDD) develops in the spine as vertebral discs degenerate and osseous excrescences or outgrowths naturally form to restabilize unstable segments of the spine. These osseous excrescences, or osteophytes, may progress or stabilize in size as the spine reaches a new equilibrium point. We have previously created a CAD system that detects DDD. This paper presents a new system to determine the severity of DDD of individual vertebral levels. This will be useful to monitor the progress of developing DDD, as rapid growth may indicate that there is a greater stabilization problem that should be addressed. The existing DDD CAD system extracts the spine from CT images and segments the cortical shell of individual levels with a dual-surface model. The cortical shell is unwrapped, and is analyzed to detect the hyperdense regions of DDD. Three radiologists scored the severity of DDD of each disc space of 46 CT scans. Radiologists' scores and features generated from CAD detections were used to train a random forest classifier. The classifier then assessed the severity of DDD at each vertebral disc level. The agreement between the computer severity score and the average radiologist's score had a quadratic weighted Cohen's kappa of 0.64.

Gender classification in low-resolution surveillance video: in-depth comparison of random forests and SVMs

NASA Astrophysics Data System (ADS)

Geelen, Christopher D.; Wijnhoven, Rob G. J.; Dubbelman, Gijs; de With, Peter H. N.

2015-03-01

This research considers gender classification in surveillance environments, typically involving low-resolution images and a large amount of viewpoint variations and occlusions. Gender classification is inherently difficult due to the large intra-class variation and interclass correlation. We have developed a gender classification system, which is successfully evaluated on two novel datasets, which realistically consider the above conditions, typical for surveillance. The system reaches a mean accuracy of up to 90% and approaches our human baseline of 92.6%, proving a high-quality gender classification system. We also present an in-depth discussion of the fundamental differences between SVM and RF classifiers. We conclude that balancing the degree of randomization in any classifier is required for the highest classification accuracy. For our problem, an RF-SVM hybrid classifier exploiting the combination of HSV and LBP features results in the highest classification accuracy of 89.9 0.2%, while classification computation time is negligible compared to the detection time of pedestrians.

Probabilistic Multi-Person Tracking Using Dynamic Bayes Networks

NASA Astrophysics Data System (ADS)

Klinger, T.; Rottensteiner, F.; Heipke, C.

2015-08-01

Tracking-by-detection is a widely used practice in recent tracking systems. These usually rely on independent single frame detections that are handled as observations in a recursive estimation framework. If these observations are imprecise the generated trajectory is prone to be updated towards a wrong position. In contrary to existing methods our novel approach uses a Dynamic Bayes Network in which the state vector of a recursive Bayes filter, as well as the location of the tracked object in the image are modelled as unknowns. These unknowns are estimated in a probabilistic framework taking into account a dynamic model, and a state-of-the-art pedestrian detector and classifier. The classifier is based on the Random Forest-algorithm and is capable of being trained incrementally so that new training samples can be incorporated at runtime. This allows the classifier to adapt to the changing appearance of a target and to unlearn outdated features. The approach is evaluated on a publicly available benchmark. The results confirm that our approach is well suited for tracking pedestrians over long distances while at the same time achieving comparatively good geometric accuracy.

A review of classification algorithms for EEG-based brain–computer interfaces: a 10 year update

NASA Astrophysics Data System (ADS)

Lotte, F.; Bougrain, L.; Cichocki, A.; Clerc, M.; Congedo, M.; Rakotomamonjy, A.; Yger, F.

2018-06-01

Objective. Most current electroencephalography (EEG)-based brain–computer interfaces (BCIs) are based on machine learning algorithms. There is a large diversity of classifier types that are used in this field, as described in our 2007 review paper. Now, approximately ten years after this review publication, many new algorithms have been developed and tested to classify EEG signals in BCIs. The time is therefore ripe for an updated review of EEG classification algorithms for BCIs. Approach. We surveyed the BCI and machine learning literature from 2007 to 2017 to identify the new classification approaches that have been investigated to design BCIs. We synthesize these studies in order to present such algorithms, to report how they were used for BCIs, what were the outcomes, and to identify their pros and cons. Main results. We found that the recently designed classification algorithms for EEG-based BCIs can be divided into four main categories: adaptive classifiers, matrix and tensor classifiers, transfer learning and deep learning, plus a few other miscellaneous classifiers. Among these, adaptive classifiers were demonstrated to be generally superior to static ones, even with unsupervised adaptation. Transfer learning can also prove useful although the benefits of transfer learning remain unpredictable. Riemannian geometry-based methods have reached state-of-the-art performances on multiple BCI problems and deserve to be explored more thoroughly, along with tensor-based methods. Shrinkage linear discriminant analysis and random forests also appear particularly useful for small training samples settings. On the other hand, deep learning methods have not yet shown convincing improvement over state-of-the-art BCI methods. Significance. This paper provides a comprehensive overview of the modern classification algorithms used in EEG-based BCIs, presents the principles of these methods and guidelines on when and how to use them. It also identifies a number of challenges to further advance EEG classification in BCI.

Detection of compensatory balance responses using wearable electromyography sensors for fall-risk assessment.

PubMed

Nouredanesh, Mina; Kukreja, Sunil L; Tung, James

2016-08-01

Loss of balance is prevalent in older adults and populations with gait and balance impairments. The present paper aims to develop a method to automatically distinguish compensatory balance responses (CBRs) from normal gait, based on activity patterns of muscles involved in maintaining balance. In this study, subjects were perturbed by lateral pushes while walking and surface electromyography (sEMG) signals were recorded from four muscles in their right leg. To extract sEMG time domain features, several filtering characteristics and segmentation approaches are examined. The performance of three classification methods, i.e., k-nearest neighbor, support vector machines, and random forests, were investigated for accurate detection of CBRs. Our results show that features extracted in the 50-200Hz band, segmented using peak sEMG amplitudes, and a random forest classifier detected CBRs with an accuracy of 92.35%. Moreover, our results support the important role of biceps femoris and rectus femoris muscles in stabilization and consequently discerning CBRs. This study contributes towards the development of wearable sensor systems to accurately and reliably monitor gait and balance control behavior in at-home settings (unsupervised conditions), over long periods of time, towards personalized fall risk assessment tools.

Appendix 2: Risk-based framework and risk case studies. Risk-based framework for evaluating changes in response thresholds and vulnerabilities.

Treesearch

Dennis S. Ojima; Louis R. Iverson; Brent L. Sohngen

2012-01-01

Alaskan forests cover one-third of the state’s 52 million ha of land (Parson et al. 2001), and are regionally and globally significant. Ninety percent of Alaskan forests are classified as boreal, representing 4 percent of the world’s boreal forests, and are located throughout interior and south-central Alaska (fig. A1-1). The remaining 10 percent of Alaskan forests are...

A Step Towards EEG-based Brain Computer Interface for Autism Intervention*

PubMed Central

Fan, Jing; Wade, Joshua W.; Bian, Dayi; Key, Alexandra P.; Warren, Zachary E.; Mion, Lorraine C.; Sarkar, Nilanjan

2017-01-01

Autism Spectrum Disorder (ASD) is a prevalent and costly neurodevelopmental disorder. Individuals with ASD often have deficits in social communication skills as well as adaptive behavior skills related to daily activities. We have recently designed a novel virtual reality (VR) based driving simulator for driving skill training for individuals with ASD. In this paper, we explored the feasibility of detecting engagement level, emotional states, and mental workload during VR-based driving using EEG as a first step towards a potential EEG-based Brain Computer Interface (BCI) for assisting autism intervention. We used spectral features of EEG signals from a 14-channel EEG neuroheadset, together with therapist ratings of behavioral engagement, enjoyment, frustration, boredom, and difficulty to train a group of classification models. Seven classification methods were applied and compared including Bayes network, naïve Bayes, Support Vector Machine (SVM), multilayer perceptron, K-nearest neighbors (KNN), random forest, and J48. The classification results were promising, with over 80% accuracy in classifying engagement and mental workload, and over 75% accuracy in classifying emotional states. Such results may lead to an adaptive closed-loop VR-based skill training system for use in autism intervention. PMID:26737113

Remote sensing based detection of forested wetlands: An evaluation of LiDAR, aerial imagery, and their data fusion

NASA Astrophysics Data System (ADS)

Suiter, Ashley Elizabeth

Multi-spectral imagery provides a robust and low-cost dataset for assessing wetland extent and quality over broad regions and is frequently used for wetland inventories. However in forested wetlands, hydrology is obscured by tree canopy making it difficult to detect with multi-spectral imagery alone. Because of this, classification of forested wetlands often includes greater errors than that of other wetlands types. Elevation and terrain derivatives have been shown to be useful for modelling wetland hydrology. But, few studies have addressed the use of LiDAR intensity data detecting hydrology in forested wetlands. Due the tendency of LiDAR signal to be attenuated by water, this research proposed the fusion of LiDAR intensity data with LiDAR elevation, terrain data, and aerial imagery, for the detection of forested wetland hydrology. We examined the utility of LiDAR intensity data and determined whether the fusion of Lidar derived data with multispectral imagery increased the accuracy of forested wetland classification compared with a classification performed with only multi-spectral image. Four classifications were performed: Classification A -- All Imagery, Classification B -- All LiDAR, Classification C -- LiDAR without Intensity, and Classification D -- Fusion of All Data. These classifications were performed using random forest and each resulted in a 3-foot resolution thematic raster of forested upland and forested wetland locations in Vermilion County, Illinois. The accuracies of these classifications were compared using Kappa Coefficient of Agreement. Importance statistics produced within the random forest classifier were evaluated in order to understand the contribution of individual datasets. Classification D, which used the fusion of LiDAR and multi-spectral imagery as input variables, had moderate to strong agreement between reference data and classification results. It was found that Classification A performed using all the LiDAR data and its derivatives (intensity, elevation, slope, aspect, curvatures, and Topographic Wetness Index) was the most accurate classification with Kappa: 78.04%, indicating moderate to strong agreement. However, Classification C, performed with LiDAR derivative without intensity data had less agreement than would be expected by chance, indicating that LiDAR contributed significantly to the accuracy of Classification B.

Validating predictions from climate envelope models

USGS Publications Warehouse

Watling, J.; Bucklin, D.; Speroterra, C.; Brandt, L.; Cabal, C.; Romañach, Stephanie S.; Mazzotti, Frank J.

2013-01-01

Climate envelope models are a potentially important conservation tool, but their ability to accurately forecast species’ distributional shifts using independent survey data has not been fully evaluated. We created climate envelope models for 12 species of North American breeding birds previously shown to have experienced poleward range shifts. For each species, we evaluated three different approaches to climate envelope modeling that differed in the way they treated climate-induced range expansion and contraction, using random forests and maximum entropy modeling algorithms. All models were calibrated using occurrence data from 1967–1971 (t1) and evaluated using occurrence data from 1998–2002 (t2). Model sensitivity (the ability to correctly classify species presences) was greater using the maximum entropy algorithm than the random forest algorithm. Although sensitivity did not differ significantly among approaches, for many species, sensitivity was maximized using a hybrid approach that assumed range expansion, but not contraction, in t2. Species for which the hybrid approach resulted in the greatest improvement in sensitivity have been reported from more land cover types than species for which there was little difference in sensitivity between hybrid and dynamic approaches, suggesting that habitat generalists may be buffered somewhat against climate-induced range contractions. Specificity (the ability to correctly classify species absences) was maximized using the random forest algorithm and was lowest using the hybrid approach. Overall, our results suggest cautious optimism for the use of climate envelope models to forecast range shifts, but also underscore the importance of considering non-climate drivers of species range limits. The use of alternative climate envelope models that make different assumptions about range expansion and contraction is a new and potentially useful way to help inform our understanding of climate change effects on species.

Automatic Seismic-Event Classification with Convolutional Neural Networks.

NASA Astrophysics Data System (ADS)

Bueno Rodriguez, A.; Titos Luzón, M.; Garcia Martinez, L.; Benitez, C.; Ibáñez, J. M.

2017-12-01

Active volcanoes exhibit a wide range of seismic signals, providing vast amounts of unlabelled volcano-seismic data that can be analyzed through the lens of artificial intelligence. However, obtaining high-quality labelled data is time-consuming and expensive. Deep neural networks can process data in their raw form, compute high-level features and provide a better representation of the input data distribution. These systems can be deployed to classify seismic data at scale, enhance current early-warning systems and build extensive seismic catalogs. In this research, we aim to classify spectrograms from seven different seismic events registered at "Volcán de Fuego" (Colima, Mexico), during four eruptive periods. Our approach is based on convolutional neural networks (CNNs), a sub-type of deep neural networks that can exploit grid structure from the data. Volcano-seismic signals can be mapped into a grid-like structure using the spectrogram: a representation of the temporal evolution in terms of time and frequency. Spectrograms were computed from the data using Hamming windows with 4 seconds length, 2.5 seconds overlapping and 128 points FFT resolution. Results are compared to deep neural networks, random forest and SVMs. Experiments show that CNNs can exploit temporal and frequency information, attaining a classification accuracy of 93%, similar to deep networks 91% but outperforming SVM and random forest. These results empirically show that CNNs are powerful models to classify a wide range of volcano-seismic signals, and achieve good generalization. Furthermore, volcano-seismic spectrograms contains useful discriminative information for the CNN, as higher layers of the network combine high-level features computed for each frequency band, helping to detect simultaneous events in time. Being at the intersection of deep learning and geophysics, this research enables future studies of how CNNs can be used in volcano monitoring to accurately determine the detection and location of seismic events.

Gis-Based Multi-Criteria Decision Analysis for Forest Fire Risk Mapping

NASA Astrophysics Data System (ADS)

Akay, A. E.; Erdoğan, A.

2017-11-01

The forested areas along the coastal zone of the Mediterranean region in Turkey are classified as first-degree fire sensitive areas. Forest fires are major environmental disaster that affects the sustainability of forest ecosystems. Besides, forest fires result in important economic losses and even threaten human lives. Thus, it is critical to determine the forested areas with fire risks and thereby minimize the damages on forest resources by taking necessary precaution measures in these areas. The risk of forest fire can be assessed based on various factors such as forest vegetation structures (tree species, crown closure, tree stage), topographic features (slope and aspect), and climatic parameters (temperature, wind). In this study, GIS-based Multi-Criteria Decision Analysis (MCDA) method was used to generate forest fire risk map. The study was implemented in the forested areas within Yayla Forest Enterprise Chiefs at Dursunbey Forest Enterprise Directorate which is classified as first degree fire sensitive area. In the solution process, "extAhp 2.0" plug-in running Analytic Hierarchy Process (AHP) method in ArcGIS 10.4.1 was used to categorize study area under five fire risk classes: extreme risk, high risk, moderate risk, and low risk. The results indicated that 23.81 % of the area was of extreme risk, while 25.81 % was of high risk. The result indicated that the most effective criterion was tree species, followed by tree stages. The aspect had the least effective criterion on forest fire risk. It was revealed that GIS techniques integrated with MCDA methods are effective tools to quickly estimate forest fire risk at low cost. The integration of these factors into GIS can be very useful to determine forested areas with high fire risk and also to plan forestry management after fire.

Approximating prediction uncertainty for random forest regression models

Treesearch

John W. Coulston; Christine E. Blinn; Valerie A. Thomas; Randolph H. Wynne

2016-01-01

Machine learning approaches such as random forest have increased for the spatial modeling and mapping of continuous variables. Random forest is a non-parametric ensemble approach, and unlike traditional regression approaches there is no direct quantification of prediction error. Understanding prediction uncertainty is important when using model-based continuous maps as...

Exploiting machine learning algorithms for tree species classification in a semiarid woodland using RapidEye image

NASA Astrophysics Data System (ADS)

Adelabu, Samuel; Mutanga, Onisimo; Adam, Elhadi; Cho, Moses Azong

2013-01-01

Classification of different tree species in semiarid areas can be challenging as a result of the change in leaf structure and orientation due to soil moisture constraints. Tree species mapping is, however, a key parameter for forest management in semiarid environments. In this study, we examined the suitability of 5-band RapidEye satellite data for the classification of five tree species in mopane woodland of Botswana using machine leaning algorithms with limited training samples.We performed classification using random forest (RF) and support vector machines (SVM) based on EnMap box. The overall accuracies for classifying the five tree species was 88.75 and 85% for both SVM and RF, respectively. We also demonstrated that the new red-edge band in the RapidEye sensor has the potential for classifying tree species in semiarid environments when integrated with other standard bands. Similarly, we observed that where there are limited training samples, SVM is preferred over RF. Finally, we demonstrated that the two accuracy measures of quantity and allocation disagreement are simpler and more helpful for the vast majority of remote sensing classification process than the kappa coefficient. Overall, high species classification can be achieved using strategically located RapidEye bands integrated with advanced processing algorithms.

Detecting understory plant invasion in urban forests using LiDAR

NASA Astrophysics Data System (ADS)

Singh, Kunwar K.; Davis, Amy J.; Meentemeyer, Ross K.

2015-06-01

Light detection and ranging (LiDAR) data are increasingly used to measure structural characteristics of urban forests but are rarely used to detect the growing problem of exotic understory plant invaders. We explored the merits of using LiDAR-derived metrics alone and through integration with spectral data to detect the spatial distribution of the exotic understory plant Ligustrum sinense, a rapidly spreading invader in the urbanizing region of Charlotte, North Carolina, USA. We analyzed regional-scale L. sinense occurrence data collected over the course of three years with LiDAR-derived metrics of forest structure that were categorized into the following groups: overstory, understory, topography, and overall vegetation characteristics, and IKONOS spectral features - optical. Using random forest (RF) and logistic regression (LR) classifiers, we assessed the relative contributions of LiDAR and IKONOS derived variables to the detection of L. sinense. We compared the top performing models developed for a smaller, nested experimental extent using RF and LR classifiers, and used the best overall model to produce a predictive map of the spatial distribution of L. sinense across our country-wide study extent. RF classification of LiDAR-derived topography metrics produced the highest mapping accuracy estimates, outperforming IKONOS data by 17.5% and the integration of LiDAR and IKONOS data by 5.3%. The top performing model from the RF classifier produced the highest kappa of 64.8%, improving on the parsimonious LR model kappa by 31.1% with a moderate gain of 6.2% over the county extent model. Our results demonstrate the superiority of LiDAR-derived metrics over spectral data and fusion of LiDAR and spectral data for accurately mapping the spatial distribution of the forest understory invader L. sinense.

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.

Deep convolutional neural network training enrichment using multi-view object-based analysis of Unmanned Aerial systems imagery for wetlands classification

NASA Astrophysics Data System (ADS)

Liu, Tao; Abd-Elrahman, Amr

2018-05-01

Deep convolutional neural network (DCNN) requires massive training datasets to trigger its image classification power, while collecting training samples for remote sensing application is usually an expensive process. When DCNN is simply implemented with traditional object-based image analysis (OBIA) for classification of Unmanned Aerial systems (UAS) orthoimage, its power may be undermined if the number training samples is relatively small. This research aims to develop a novel OBIA classification approach that can take advantage of DCNN by enriching the training dataset automatically using multi-view data. Specifically, this study introduces a Multi-View Object-based classification using Deep convolutional neural network (MODe) method to process UAS images for land cover classification. MODe conducts the classification on multi-view UAS images instead of directly on the orthoimage, and gets the final results via a voting procedure. 10-fold cross validation results show the mean overall classification accuracy increasing substantially from 65.32%, when DCNN was applied on the orthoimage to 82.08% achieved when MODe was implemented. This study also compared the performances of the support vector machine (SVM) and random forest (RF) classifiers with DCNN under traditional OBIA and the proposed multi-view OBIA frameworks. The results indicate that the advantage of DCNN over traditional classifiers in terms of accuracy is more obvious when these classifiers were applied with the proposed multi-view OBIA framework than when these classifiers were applied within the traditional OBIA framework.

A comprehensive simulation study on classification of RNA-Seq data.

PubMed

Zararsız, Gökmen; Goksuluk, Dincer; Korkmaz, Selcuk; Eldem, Vahap; Zararsiz, Gozde Erturk; Duru, Izzet Parug; Ozturk, Ahmet

2017-01-01

RNA sequencing (RNA-Seq) is a powerful technique for the gene-expression profiling of organisms that uses the capabilities of next-generation sequencing technologies. Developing gene-expression-based classification algorithms is an emerging powerful method for diagnosis, disease classification and monitoring at molecular level, as well as providing potential markers of diseases. Most of the statistical methods proposed for the classification of gene-expression data are either based on a continuous scale (eg. microarray data) or require a normal distribution assumption. Hence, these methods cannot be directly applied to RNA-Seq data since they violate both data structure and distributional assumptions. However, it is possible to apply these algorithms with appropriate modifications to RNA-Seq data. One way is to develop count-based classifiers, such as Poisson linear discriminant analysis and negative binomial linear discriminant analysis. Another way is to bring the data closer to microarrays and apply microarray-based classifiers. In this study, we compared several classifiers including PLDA with and without power transformation, NBLDA, single SVM, bagging SVM (bagSVM), classification and regression trees (CART), and random forests (RF). We also examined the effect of several parameters such as overdispersion, sample size, number of genes, number of classes, differential-expression rate, and the transformation method on model performances. A comprehensive simulation study is conducted and the results are compared with the results of two miRNA and two mRNA experimental datasets. The results revealed that increasing the sample size, differential-expression rate and decreasing the dispersion parameter and number of groups lead to an increase in classification accuracy. Similar with differential-expression studies, the classification of RNA-Seq data requires careful attention when handling data overdispersion. We conclude that, as a count-based classifier, the power transformed PLDA and, as a microarray-based classifier, vst or rlog transformed RF and SVM classifiers may be a good choice for classification. An R/BIOCONDUCTOR package, MLSeq, is freely available at https://www.bioconductor.org/packages/release/bioc/html/MLSeq.html.

Machine learning algorithms for the creation of clinical healthcare enterprise systems

NASA Astrophysics Data System (ADS)

Mandal, Indrajit

2017-10-01

Clinical recommender systems are increasingly becoming popular for improving modern healthcare systems. Enterprise systems are persuasively used for creating effective nurse care plans to provide nurse training, clinical recommendations and clinical quality control. A novel design of a reliable clinical recommender system based on multiple classifier system (MCS) is implemented. A hybrid machine learning (ML) ensemble based on random subspace method and random forest is presented. The performance accuracy and robustness of proposed enterprise architecture are quantitatively estimated to be above 99% and 97%, respectively (above 95% confidence interval). The study then extends to experimental analysis of the clinical recommender system with respect to the noisy data environment. The ranking of items in nurse care plan is demonstrated using machine learning algorithms (MLAs) to overcome the drawback of the traditional association rule method. The promising experimental results are compared against the sate-of-the-art approaches to highlight the advancement in recommendation technology. The proposed recommender system is experimentally validated using five benchmark clinical data to reinforce the research findings.

Computer assisted optical biopsy for colorectal polyps

NASA Astrophysics Data System (ADS)

Navarro-Avila, Fernando J.; Saint-Hill-Febles, Yadira; Renner, Janis; Klare, Peter; von Delius, Stefan; Navab, Nassir; Mateus, Diana

2017-03-01

We propose a method for computer-assisted optical biopsy for colorectal polyps, with the final goal of assisting the medical expert during the colonoscopy. In particular, we target the problem of automatic classification of polyp images in two classes: adenomatous vs non-adenoma. Our approach is based on recent advancements in convolutional neural networks (CNN) for image representation. In the paper, we describe and compare four different methodologies to address the binary classification task: a baseline with classical features and a Random Forest classifier, two methods based on features obtained from a pre-trained network, and finally, the end-to-end training of a CNN. With the pre-trained network, we show the feasibility of transferring a feature extraction mechanism trained on millions of natural images, to the task of classifying adenomatous polyps. We then demonstrate further performance improvements when training the CNN for our specific classification task. In our study, 776 polyp images were acquired and histologically analyzed after polyp resection. We report a performance increase of the CNN-based approaches with respect to both, the conventional engineered features and to a state-of-the-art method based on videos and 3D shape features.

Fragmentation of Continental United States Forests

Treesearch

Kurt H. Riitters; James D. Wickham; Robert V. O' Neill; K. Bruce Jones; Elizabeth R. Smith; John W. Coulston; Timothy G. Wade; Jonathan H. Smith

2002-01-01

We report a multiple-scale analysis of forest fragmentation based on 30-m (0.09 ha pixel-1) land- cover maps for the conterminous United States. Each 0.09-ha unit of forest was classified according to fragmentation indexes measured within the surrounding landscape, for five landscape sizes including 2.25, 7.29, 65.61, 590.49, and 5314.41 ha....

Forest Species Diversity in Upper Elevation Hardwood Forests in the Southern Appalachian Mountains

Treesearch

Katherine J. Elliott; Deidre Hewitt

1997-01-01

Overstory, shrub-layer, and herb-layer flora composition and abundance patterns in eleven forest sites were studied to evaluate species diversity and richness before implementing three types of harvest treat- ments. The sites were within the Wine Spring Creek Watershed and were classified as high elevation, dry, Quercus rubra-Rhododendron calendulaceum based on...

Time Series of Images to Improve Tree Species Classification

NASA Astrophysics Data System (ADS)

Miyoshi, G. T.; Imai, N. N.; de Moraes, M. V. A.; Tommaselli, A. M. G.; Näsi, R.

2017-10-01

Tree species classification provides valuable information to forest monitoring and management. The high floristic variation of the tree species appears as a challenging issue in the tree species classification because the vegetation characteristics changes according to the season. To help to monitor this complex environment, the imaging spectroscopy has been largely applied since the development of miniaturized sensors attached to Unmanned Aerial Vehicles (UAV). Considering the seasonal changes in forests and the higher spectral and spatial resolution acquired with sensors attached to UAV, we present the use of time series of images to classify four tree species. The study area is an Atlantic Forest area located in the western part of São Paulo State. Images were acquired in August 2015 and August 2016, generating three data sets of images: only with the image spectra of 2015; only with the image spectra of 2016; with the layer stacking of images from 2015 and 2016. Four tree species were classified using Spectral angle mapper (SAM), Spectral information divergence (SID) and Random Forest (RF). The results showed that SAM and SID caused an overfitting of the data whereas RF showed better results and the use of the layer stacking improved the classification achieving a kappa coefficient of 18.26 %.

A minimum spanning forest based classification method for dedicated breast CT images

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

Pike, Robert; Sechopoulos, Ioannis; Fei, Baowei, E-mail: bfei@emory.edu

Purpose: To develop and test an automated algorithm to classify different types of tissue in dedicated breast CT images. Methods: Images of a single breast of five different patients were acquired with a dedicated breast CT clinical prototype. The breast CT images were processed by a multiscale bilateral filter to reduce noise while keeping edge information and were corrected to overcome cupping artifacts. As skin and glandular tissue have similar CT values on breast CT images, morphologic processing is used to identify the skin based on its position information. A support vector machine (SVM) is trained and the resulting modelmore » used to create a pixelwise classification map of fat and glandular tissue. By combining the results of the skin mask with the SVM results, the breast tissue is classified as skin, fat, and glandular tissue. This map is then used to identify markers for a minimum spanning forest that is grown to segment the image using spatial and intensity information. To evaluate the authors’ classification method, they use DICE overlap ratios to compare the results of the automated classification to those obtained by manual segmentation on five patient images. Results: Comparison between the automatic and the manual segmentation shows that the minimum spanning forest based classification method was able to successfully classify dedicated breast CT image with average DICE ratios of 96.9%, 89.8%, and 89.5% for fat, glandular, and skin tissue, respectively. Conclusions: A 2D minimum spanning forest based classification method was proposed and evaluated for classifying the fat, skin, and glandular tissue in dedicated breast CT images. The classification method can be used for dense breast tissue quantification, radiation dose assessment, and other applications in breast imaging.« less

Forest tree species clssification based on airborne hyper-spectral imagery

NASA Astrophysics Data System (ADS)

Dian, Yuanyong; Li, Zengyuan; Pang, Yong

2013-10-01

Forest precision classification products were the basic data for surveying of forest resource, updating forest subplot information, logging and design of forest. However, due to the diversity of stand structure, complexity of the forest growth environment, it's difficult to discriminate forest tree species using multi-spectral image. The airborne hyperspectral images can achieve the high spatial and spectral resolution imagery of forest canopy, so it will good for tree species level classification. The aim of this paper was to test the effective of combining spatial and spectral features in airborne hyper-spectral image classification. The CASI hyper spectral image data were acquired from Liangshui natural reserves area. Firstly, we use the MNF (minimum noise fraction) transform method for to reduce the hyperspectral image dimensionality and highlighting variation. And secondly, we use the grey level co-occurrence matrix (GLCM) to extract the texture features of forest tree canopy from the hyper-spectral image, and thirdly we fused the texture and the spectral features of forest canopy to classify the trees species using support vector machine (SVM) with different kernel functions. The results showed that when using the SVM classifier, MNF and texture-based features combined with linear kernel function can achieve the best overall accuracy which was 85.92%. It was also confirm that combine the spatial and spectral information can improve the accuracy of tree species classification.

A framework for mapping tree species combining hyperspectral and LiDAR data: Role of selected classifiers and sensor across three spatial scales

NASA Astrophysics Data System (ADS)

Ghosh, Aniruddha; Fassnacht, Fabian Ewald; Joshi, P. K.; Koch, Barbara

2014-02-01

Knowledge of tree species distribution is important worldwide for sustainable forest management and resource evaluation. The accuracy and information content of species maps produced using remote sensing images vary with scale, sensor (optical, microwave, LiDAR), classification algorithm, verification design and natural conditions like tree age, forest structure and density. Imaging spectroscopy reduces the inaccuracies making use of the detailed spectral response. However, the scale effect still has a strong influence and cannot be neglected. This study aims to bridge the knowledge gap in understanding the scale effect in imaging spectroscopy when moving from 4 to 30 m pixel size for tree species mapping, keeping in mind that most current and future hyperspectral satellite based sensors work with spatial resolution around 30 m or more. Two airborne (HyMAP) and one spaceborne (Hyperion) imaging spectroscopy dataset with pixel sizes of 4, 8 and 30 m, respectively were available to examine the effect of scale over a central European forest. The forest under examination is a typical managed forest with relatively homogenous stands featuring mostly two canopy layers. Normalized digital surface model (nDSM) derived from LiDAR data was used additionally to examine the effect of height information in tree species mapping. Six different sets of predictor variables (reflectance value of all bands, selected components of a Minimum Noise Fraction (MNF), Vegetation Indices (VI) and each of these sets combined with LiDAR derived height) were explored at each scale. Supervised kernel based (Support Vector Machines) and ensemble based (Random Forest) machine learning algorithms were applied on the dataset to investigate the effect of the classifier. Iterative bootstrap-validation with 100 iterations was performed for classification model building and testing for all the trials. For scale, analysis of overall classification accuracy and kappa values indicated that 8 m spatial resolution (reaching kappa values of over 0.83) slightly outperformed the results obtained from 4 m for the study area and five tree species under examination. The 30 m resolution Hyperion image produced sound results (kappa values of over 0.70), which in some areas of the test site were comparable with the higher spatial resolution imagery when qualitatively assessing the map outputs. Considering input predictor sets, MNF bands performed best at 4 and 8 m resolution. Optical bands were found to be best for 30 m spatial resolution. Classification with MNF as input predictors produced better visual appearance of tree species patches when compared with reference maps. Based on the analysis, it was concluded that there is no significant effect of height information on tree species classification accuracies for the present framework and study area. Furthermore, in the examined cases there was no single best choice among the two classifiers across scales and predictors. It can be concluded that tree species mapping from imaging spectroscopy for forest sites comparable to the one under investigation is possible with reliable accuracies not only from airborne but also from spaceborne imaging spectroscopy datasets.

Discovering disease-associated genes in weighted protein-protein interaction networks

NASA Astrophysics Data System (ADS)

Cui, Ying; Cai, Meng; Stanley, H. Eugene

2018-04-01

Although there have been many network-based attempts to discover disease-associated genes, most of them have not taken edge weight - which quantifies their relative strength - into consideration. We use connection weights in a protein-protein interaction (PPI) network to locate disease-related genes. We analyze the topological properties of both weighted and unweighted PPI networks and design an improved random forest classifier to distinguish disease genes from non-disease genes. We use a cross-validation test to confirm that weighted networks are better able to discover disease-associated genes than unweighted networks, which indicates that including link weight in the analysis of network properties provides a better model of complex genotype-phenotype associations.

Mississippi's forests, 2006

Treesearch

Sonja N. Oswalt; Tony G. Johnson; John W. Coulston; Christopher M. Oswalt

2009-01-01

Forest land covers 19.6 million acres in Mississippi, or about 65 percent of the land area. The majority of forests are classed as timberland. One hundred and thirty-seven tree species were measured on Mississippi forests in the 2006 inventory. Thirty six percent of Mississippi's forest land is classified as loblolly-shortleaf pine forest, 27 percent is classified...

Applying data fusion techniques for benthic habitat mapping and monitoring in a coral reef ecosystem

NASA Astrophysics Data System (ADS)

Zhang, Caiyun

2015-06-01

Accurate mapping and effective monitoring of benthic habitat in the Florida Keys are critical in developing management strategies for this valuable coral reef ecosystem. For this study, a framework was designed for automated benthic habitat mapping by combining multiple data sources (hyperspectral, aerial photography, and bathymetry data) and four contemporary imagery processing techniques (data fusion, Object-based Image Analysis (OBIA), machine learning, and ensemble analysis). In the framework, 1-m digital aerial photograph was first merged with 17-m hyperspectral imagery and 10-m bathymetry data using a pixel/feature-level fusion strategy. The fused dataset was then preclassified by three machine learning algorithms (Random Forest, Support Vector Machines, and k-Nearest Neighbor). Final object-based habitat maps were produced through ensemble analysis of outcomes from three classifiers. The framework was tested for classifying a group-level (3-class) and code-level (9-class) habitats in a portion of the Florida Keys. Informative and accurate habitat maps were achieved with an overall accuracy of 88.5% and 83.5% for the group-level and code-level classifications, respectively.

Recognizing stationary and locomotion activities using combinational of spectral analysis with statistical descriptors features

NASA Astrophysics Data System (ADS)

Zainudin, M. N. Shah; Sulaiman, Md Nasir; Mustapha, Norwati; Perumal, Thinagaran

2017-10-01

Prior knowledge in pervasive computing recently garnered a lot of attention due to its high demand in various application domains. Human activity recognition (HAR) considered as the applications that are widely explored by the expertise that provides valuable information to the human. Accelerometer sensor-based approach is utilized as devices to undergo the research in HAR since their small in size and this sensor already build-in in the various type of smartphones. However, the existence of high inter-class similarities among the class tends to degrade the recognition performance. Hence, this work presents the method for activity recognition using our proposed features from combinational of spectral analysis with statistical descriptors that able to tackle the issue of differentiating stationary and locomotion activities. The noise signal is filtered using Fourier Transform before it will be extracted using two different groups of features, spectral frequency analysis, and statistical descriptors. Extracted signal later will be classified using random forest ensemble classifier models. The recognition results show the good accuracy performance for stationary and locomotion activities based on USC HAD datasets.

Keystroke Dynamics-Based Credential Hardening Systems

NASA Astrophysics Data System (ADS)

Bartlow, Nick; Cukic, Bojan

abstract Keystroke dynamics are becoming a well-known method for strengthening username- and password-based credential sets. The familiarity and ease of use of these traditional authentication schemes combined with the increased trustworthiness associated with biometrics makes them prime candidates for application in many web-based scenarios. Our keystroke dynamics system uses Breiman’s random forests algorithm to classify keystroke input sequences as genuine or imposter. The system is capable of operating at various points on a traditional ROC curve depending on application-specific security needs. As a username/password authentication scheme, our approach decreases the system penetration rate associated with compromised passwords up to 99.15%. Beyond presenting results demonstrating the credential hardening effect of our scheme, we look into the notion that a user’s familiarity to components of a credential set can non-trivially impact error rates.

Hyperspectral detection of a subsurface CO2 leak in the presence of water stressed vegetation.

PubMed

Bellante, Gabriel J; Powell, Scott L; Lawrence, Rick L; Repasky, Kevin S; Dougher, Tracy

2014-01-01

Remote sensing of vegetation stress has been posed as a possible large area monitoring tool for surface CO2 leakage from geologic carbon sequestration (GCS) sites since vegetation is adversely affected by elevated CO2 levels in soil. However, the extent to which remote sensing could be used for CO2 leak detection depends on the spectral separability of the plant stress signal caused by various factors, including elevated soil CO2 and water stress. This distinction is crucial to determining the seasonality and appropriateness of remote GCS site monitoring. A greenhouse experiment tested the degree to which plants stressed by elevated soil CO2 could be distinguished from plants that were water stressed. A randomized block design assigned Alfalfa plants (Medicago sativa) to one of four possible treatment groups: 1) a CO2 injection group; 2) a water stress group; 3) an interaction group that was subjected to both water stress and CO2 injection; or 4) a group that received adequate water and no CO2 injection. Single date classification trees were developed to identify individual spectral bands that were significant in distinguishing between CO2 and water stress agents, in addition to a random forest classifier that was used to further understand and validate predictive accuracies. Overall peak classification accuracy was 90% (Kappa of 0.87) for the classification tree analysis and 83% (Kappa of 0.77) for the random forest classifier, demonstrating that vegetation stressed from an underground CO2 leak could be accurately discerned from healthy vegetation and areas of co-occurring water stressed vegetation at certain times. Plants appear to hit a stress threshold, however, that would render detection of a CO2 leak unlikely during severe drought conditions. Our findings suggest that early detection of a CO2 leak with an aerial or ground-based hyperspectral imaging system is possible and could be an important GCS monitoring tool.

Hyperspectral Detection of a Subsurface CO2 Leak in the Presence of Water Stressed Vegetation

PubMed Central

Bellante, Gabriel J.; Powell, Scott L.; Lawrence, Rick L.; Repasky, Kevin S.; Dougher, Tracy

2014-01-01

Remote sensing of vegetation stress has been posed as a possible large area monitoring tool for surface CO2 leakage from geologic carbon sequestration (GCS) sites since vegetation is adversely affected by elevated CO2 levels in soil. However, the extent to which remote sensing could be used for CO2 leak detection depends on the spectral separability of the plant stress signal caused by various factors, including elevated soil CO2 and water stress. This distinction is crucial to determining the seasonality and appropriateness of remote GCS site monitoring. A greenhouse experiment tested the degree to which plants stressed by elevated soil CO2 could be distinguished from plants that were water stressed. A randomized block design assigned Alfalfa plants (Medicago sativa) to one of four possible treatment groups: 1) a CO2 injection group; 2) a water stress group; 3) an interaction group that was subjected to both water stress and CO2 injection; or 4) a group that received adequate water and no CO2 injection. Single date classification trees were developed to identify individual spectral bands that were significant in distinguishing between CO2 and water stress agents, in addition to a random forest classifier that was used to further understand and validate predictive accuracies. Overall peak classification accuracy was 90% (Kappa of 0.87) for the classification tree analysis and 83% (Kappa of 0.77) for the random forest classifier, demonstrating that vegetation stressed from an underground CO2 leak could be accurately discerned from healthy vegetation and areas of co-occurring water stressed vegetation at certain times. Plants appear to hit a stress threshold, however, that would render detection of a CO2 leak unlikely during severe drought conditions. Our findings suggest that early detection of a CO2 leak with an aerial or ground-based hyperspectral imaging system is possible and could be an important GCS monitoring tool. PMID:25330232

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.

Mapping Fuels on the Okanogan and Wenatchee National Forests

Treesearch

Crystal L. Raymond; Lara-Karena B. Kellogg; Donald McKenzie

2006-01-01

Resource managers need spatially explicit fuels data to manage fire hazard and evaluate the ecological effects of wildland fires and fuel treatments. For this study, fuels were mapped on the Okanogan and Wenatchee National Forests (OWNF) using a rule-based method and the Fuels Characteristic Classification System (FCCS). The FCCS classifies fuels based on their...

The influence of negative training set size on machine learning-based virtual screening.

PubMed

Kurczab, Rafał; Smusz, Sabina; Bojarski, Andrzej J

2014-01-01

The paper presents a thorough analysis of the influence of the number of negative training examples on the performance of machine learning methods. The impact of this rather neglected aspect of machine learning methods application was examined for sets containing a fixed number of positive and a varying number of negative examples randomly selected from the ZINC database. An increase in the ratio of positive to negative training instances was found to greatly influence most of the investigated evaluating parameters of ML methods in simulated virtual screening experiments. In a majority of cases, substantial increases in precision and MCC were observed in conjunction with some decreases in hit recall. The analysis of dynamics of those variations let us recommend an optimal composition of training data. The study was performed on several protein targets, 5 machine learning algorithms (SMO, Naïve Bayes, Ibk, J48 and Random Forest) and 2 types of molecular fingerprints (MACCS and CDK FP). The most effective classification was provided by the combination of CDK FP with SMO or Random Forest algorithms. The Naïve Bayes models appeared to be hardly sensitive to changes in the number of negative instances in the training set. In conclusion, the ratio of positive to negative training instances should be taken into account during the preparation of machine learning experiments, as it might significantly influence the performance of particular classifier. What is more, the optimization of negative training set size can be applied as a boosting-like approach in machine learning-based virtual screening.

The influence of negative training set size on machine learning-based virtual screening

PubMed Central

2014-01-01

Background The paper presents a thorough analysis of the influence of the number of negative training examples on the performance of machine learning methods. Results The impact of this rather neglected aspect of machine learning methods application was examined for sets containing a fixed number of positive and a varying number of negative examples randomly selected from the ZINC database. An increase in the ratio of positive to negative training instances was found to greatly influence most of the investigated evaluating parameters of ML methods in simulated virtual screening experiments. In a majority of cases, substantial increases in precision and MCC were observed in conjunction with some decreases in hit recall. The analysis of dynamics of those variations let us recommend an optimal composition of training data. The study was performed on several protein targets, 5 machine learning algorithms (SMO, Naïve Bayes, Ibk, J48 and Random Forest) and 2 types of molecular fingerprints (MACCS and CDK FP). The most effective classification was provided by the combination of CDK FP with SMO or Random Forest algorithms. The Naïve Bayes models appeared to be hardly sensitive to changes in the number of negative instances in the training set. Conclusions In conclusion, the ratio of positive to negative training instances should be taken into account during the preparation of machine learning experiments, as it might significantly influence the performance of particular classifier. What is more, the optimization of negative training set size can be applied as a boosting-like approach in machine learning-based virtual screening. PMID:24976867

Accurate Segmentation of CT Male Pelvic Organs via Regression-based Deformable Models and Multi-task Random Forests

PubMed Central

Gao, Yaozong; Shao, Yeqin; Lian, Jun; Wang, Andrew Z.; Chen, Ronald C.

2016-01-01

Segmenting male pelvic organs from CT images is a prerequisite for prostate cancer radiotherapy. The efficacy of radiation treatment highly depends on segmentation accuracy. However, accurate segmentation of male pelvic organs is challenging due to low tissue contrast of CT images, as well as large variations of shape and appearance of the pelvic organs. Among existing segmentation methods, deformable models are the most popular, as shape prior can be easily incorporated to regularize the segmentation. Nonetheless, the sensitivity to initialization often limits their performance, especially for segmenting organs with large shape variations. In this paper, we propose a novel approach to guide deformable models, thus making them robust against arbitrary initializations. Specifically, we learn a displacement regressor, which predicts 3D displacement from any image voxel to the target organ boundary based on the local patch appearance. This regressor provides a nonlocal external force for each vertex of deformable model, thus overcoming the initialization problem suffered by the traditional deformable models. To learn a reliable displacement regressor, two strategies are particularly proposed. 1) A multi-task random forest is proposed to learn the displacement regressor jointly with the organ classifier; 2) an auto-context model is used to iteratively enforce structural information during voxel-wise prediction. Extensive experiments on 313 planning CT scans of 313 patients show that our method achieves better results than alternative classification or regression based methods, and also several other existing methods in CT pelvic organ segmentation. PMID:26800531

Accuracy and efficiency of area classifications based on tree tally

Treesearch

Michael S. Williams; Hans T. Schreuder; Raymond L. Czaplewski

2001-01-01

Inventory data are often used to estimate the area of the land base that is classified as a specific condition class. Examples include areas classified as old-growth forest, private ownership, or suitable habitat for a given species. Many inventory programs rely on classification algorithms of varying complexity to determine condition class. These algorithms can be...

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.

Mediastinal lymph node detection and station mapping on chest CT using spatial priors and random forest

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

Liu, Jiamin; Hoffman, Joanne; Zhao, Jocelyn

2016-07-15

Purpose: To develop an automated system for mediastinal lymph node detection and station mapping for chest CT. Methods: The contextual organs, trachea, lungs, and spine are first automatically identified to locate the region of interest (ROI) (mediastinum). The authors employ shape features derived from Hessian analysis, local object scale, and circular transformation that are computed per voxel in the ROI. Eight more anatomical structures are simultaneously segmented by multiatlas label fusion. Spatial priors are defined as the relative multidimensional distance vectors corresponding to each structure. Intensity, shape, and spatial prior features are integrated and parsed by a random forest classifiermore » for lymph node detection. The detected candidates are then segmented by the following curve evolution process. Texture features are computed on the segmented lymph nodes and a support vector machine committee is used for final classification. For lymph node station labeling, based on the segmentation results of the above anatomical structures, the textual definitions of mediastinal lymph node map according to the International Association for the Study of Lung Cancer are converted into patient-specific color-coded CT image, where the lymph node station can be automatically assigned for each detected node. Results: The chest CT volumes from 70 patients with 316 enlarged mediastinal lymph nodes are used for validation. For lymph node detection, their system achieves 88% sensitivity at eight false positives per patient. For lymph node station labeling, 84.5% of lymph nodes are correctly assigned to their stations. Conclusions: Multiple-channel shape, intensity, and spatial prior features aggregated by a random forest classifier improve mediastinal lymph node detection on chest CT. Using the location information of segmented anatomic structures from the multiatlas formulation enables accurate identification of lymph node stations.« less

Comparing the performance of meta-classifiers—a case study on selected imbalanced data sets relevant for prediction of liver toxicity

NASA Astrophysics Data System (ADS)

Jain, Sankalp; Kotsampasakou, Eleni; Ecker, Gerhard F.

2018-05-01

Cheminformatics datasets used in classification problems, especially those related to biological or physicochemical properties, are often imbalanced. This presents a major challenge in development of in silico prediction models, as the traditional machine learning algorithms are known to work best on balanced datasets. The class imbalance introduces a bias in the performance of these algorithms due to their preference towards the majority class. Here, we present a comparison of the performance of seven different meta-classifiers for their ability to handle imbalanced datasets, whereby Random Forest is used as base-classifier. Four different datasets that are directly (cholestasis) or indirectly (via inhibition of organic anion transporting polypeptide 1B1 and 1B3) related to liver toxicity were chosen for this purpose. The imbalance ratio in these datasets ranges between 4:1 and 20:1 for negative and positive classes, respectively. Three different sets of molecular descriptors for model development were used, and their performance was assessed in 10-fold cross-validation and on an independent validation set. Stratified bagging, MetaCost and CostSensitiveClassifier were found to be the best performing among all the methods. While MetaCost and CostSensitiveClassifier provided better sensitivity values, Stratified Bagging resulted in high balanced accuracies.

Comparing the performance of meta-classifiers—a case study on selected imbalanced data sets relevant for prediction of liver toxicity

NASA Astrophysics Data System (ADS)

Jain, Sankalp; Kotsampasakou, Eleni; Ecker, Gerhard F.

2018-04-01

Cheminformatics datasets used in classification problems, especially those related to biological or physicochemical properties, are often imbalanced. This presents a major challenge in development of in silico prediction models, as the traditional machine learning algorithms are known to work best on balanced datasets. The class imbalance introduces a bias in the performance of these algorithms due to their preference towards the majority class. Here, we present a comparison of the performance of seven different meta-classifiers for their ability to handle imbalanced datasets, whereby Random Forest is used as base-classifier. Four different datasets that are directly (cholestasis) or indirectly (via inhibition of organic anion transporting polypeptide 1B1 and 1B3) related to liver toxicity were chosen for this purpose. The imbalance ratio in these datasets ranges between 4:1 and 20:1 for negative and positive classes, respectively. Three different sets of molecular descriptors for model development were used, and their performance was assessed in 10-fold cross-validation and on an independent validation set. Stratified bagging, MetaCost and CostSensitiveClassifier were found to be the best performing among all the methods. While MetaCost and CostSensitiveClassifier provided better sensitivity values, Stratified Bagging resulted in high balanced accuracies.

Redistribution of soil nitrogen, carbon and organic matter by mechanical disturbance during whole-tree harvesting in northern hardwoods

USGS Publications Warehouse

Ryan, D.F.; Huntington, T.G.; Wayne, Martin C.

1992-01-01

To investigate whether mechanical mixing during harvesting could account for losses observed from forest floor, we measured surface disturbance on a 22 ha watershed that was whole-tree harvested. Surface soil on each 10 cm interval along 81, randomly placed transects was classified immediately after harvesting as mineral or organic, and as undisturbed, depressed, rutted, mounded, scarified, or scalped (forest floor scraped away). We quantitatively sampled these surface categories to collect soil in which preharvest forest floor might reside after harvest. Mechanically mixed mineral and organic soil horizons were readily identified. Buried forest floor under mixed mineral soil occurred in 57% of mounds with mineral surface soil. Harvesting disturbed 65% of the watershed surface and removed forest floor from 25% of the area. Mechanically mixed soil under ruts with organic or mineral surface soil, and mounds with mineral surface soil contained organic carbon and nitrogen pools significantly greater than undisturbed forest floor. Mechanical mixing into underlying mineral soil could account for the loss of forest floor observed between the preharvest condition and the second growing season after whole-tree harvesting. ?? 1992.

Examining conifer canopy structural complexity across forest ages and elevations with LiDAR data

Treesearch

Van R. Kane; Jonathan D. Bakker; Robert J. McGaughey; James A. Lutz; Rolf F. Gersonde; Jerry F. Franklin

2010-01-01

LiDAR measurements of canopy structure can be used to classify forest stands into structural stages to study spatial patterns of canopy structure, identify habitat, or plan management actions. A key assumption in this process is that differences in canopy structure based on forest age and elevation are consistent with predictions from models of stand development. Three...

New Tree-Classification System Used by the Southern Forest Inventory and Analysis Unit

Treesearch

Dennis M. May; John S. Vissage; D. Vince Few

1990-01-01

Trees at USDA Forest Service, Southern Forest Inventory and Analysis, sample locations are classified as growing stock or cull based on their ability to produce sawlogs. The old and new classification systems are compared, and the impacts of the new system on the reporting of tree volumes are illustrated with inventory data from north Alabama.

Characterization of cervigram image sharpness using multiple self-referenced measurements and random forest classifiers

NASA Astrophysics Data System (ADS)

Jaiswal, Mayoore; Horning, Matt; Hu, Liming; Ben-Or, Yau; Champlin, Cary; Wilson, Benjamin; Levitz, David

2018-02-01

Cervical cancer is the fourth most common cancer among women worldwide and is especially prevalent in low resource settings due to lack of screening and treatment options. Visual inspection with acetic acid (VIA) is a widespread and cost-effective screening method for cervical pre-cancer lesions, but accuracy depends on the experience level of the health worker. Digital cervicography, capturing images of the cervix, enables review by an off-site expert or potentially a machine learning algorithm. These reviews require images of sufficient quality. However, image quality varies greatly across users. A novel algorithm was developed to evaluate the sharpness of images captured with the MobileODT's digital cervicography device (EVA System), in order to, eventually provide feedback to the health worker. The key challenges are that the algorithm evaluates only a single image of each cervix, it needs to be robust to the variability in cervix images and fast enough to run in real time on a mobile device, and the machine learning model needs to be small enough to fit on a mobile device's memory, train on a small imbalanced dataset and run in real-time. In this paper, the focus scores of a preprocessed image and a Gaussian-blurred version of the image are calculated using established methods and used as features. A feature selection metric is proposed to select the top features which were then used in a random forest classifier to produce the final focus score. The resulting model, based on nine calculated focus scores, achieved significantly better accuracy than any single focus measure when tested on a holdout set of images. The area under the receiver operating characteristics curve was 0.9459.

Satellite inventory of Minnesota forest resources

NASA Technical Reports Server (NTRS)

Bauer, Marvin E.; Burk, Thomas E.; Ek, Alan R.; Coppin, Pol R.; Lime, Stephen D.; Walsh, Terese A.; Walters, David K.; Befort, William; Heinzen, David F.

1993-01-01

The methods and results of using Landsat Thematic Mapper (TM) data to classify and estimate the acreage of forest covertypes in northeastern Minnesota are described. Portions of six TM scenes covering five counties with a total area of 14,679 square miles were classified into six forest and five nonforest classes. The approach involved the integration of cluster sampling, image processing, and estimation. Using cluster sampling, 343 plots, each 88 acres in size, were photo interpreted and field mapped as a source of reference data for classifier training and calibration of the TM data classifications. Classification accuracies of up to 75 percent were achieved; most misclassification was between similar or related classes. An inverse method of calibration, based on the error rates obtained from the classifications of the cluster plots, was used to adjust the classification class proportions for classification errors. The resulting area estimates for total forest land in the five-county area were within 3 percent of the estimate made independently by the USDA Forest Service. Area estimates for conifer and hardwood forest types were within 0.8 and 6.0 percent respectively, of the Forest Service estimates. A trial of a second method of estimating the same classes as the Forest Service resulted in standard errors of 0.002 to 0.015. A study of the use of multidate TM data for change detection showed that forest canopy depletion, canopy increment, and no change could be identified with greater than 90 percent accuracy. The project results have been the basis for the Minnesota Department of Natural Resources and the Forest Service to define and begin to implement an annual system of forest inventory which utilizes Landsat TM data to detect changes in forest cover.

Effectively identifying compound-protein interactions by learning from positive and unlabeled examples.

PubMed

Cheng, Zhanzhan; Zhou, Shuigeng; Wang, Yang; Liu, Hui; Guan, Jihong; Chen, Yi-Ping Phoebe

2016-05-18

Prediction of compound-protein interactions (CPIs) is to find new compound-protein pairs where a protein is targeted by at least a compound, which is a crucial step in new drug design. Currently, a number of machine learning based methods have been developed to predict new CPIs in the literature. However, as there is not yet any publicly available set of validated negative CPIs, most existing machine learning based approaches use the unknown interactions (not validated CPIs) selected randomly as the negative examples to train classifiers for predicting new CPIs. Obviously, this is not quite reasonable and unavoidably impacts the CPI prediction performance. In this paper, we simply take the unknown CPIs as unlabeled examples, and propose a new method called PUCPI (the abbreviation of PU learning for Compound-Protein Interaction identification) that employs biased-SVM (Support Vector Machine) to predict CPIs using only positive and unlabeled examples. PU learning is a class of learning methods that leans from positive and unlabeled (PU) samples. To the best of our knowledge, this is the first work that identifies CPIs using only positive and unlabeled examples. We first collect known CPIs as positive examples and then randomly select compound-protein pairs not in the positive set as unlabeled examples. For each CPI/compound-protein pair, we extract protein domains as protein features and compound substructures as chemical features, then take the tensor product of the corresponding compound features and protein features as the feature vector of the CPI/compound-protein pair. After that, biased-SVM is employed to train classifiers on different datasets of CPIs and compound-protein pairs. Experiments over various datasets show that our method outperforms six typical classifiers, including random forest, L1- and L2-regularized logistic regression, naive Bayes, SVM and k-nearest neighbor (kNN), and three types of existing CPI prediction models. Source code, datasets and related documents of PUCPI are available at: http://admis.fudan.edu.cn/projects/pucpi.html.

Spot counting on fluorescence in situ hybridization in suspension images using Gaussian mixture model

NASA Astrophysics Data System (ADS)

Liu, Sijia; Sa, Ruhan; Maguire, Orla; Minderman, Hans; Chaudhary, Vipin

2015-03-01

Cytogenetic abnormalities are important diagnostic and prognostic criteria for acute myeloid leukemia (AML). A flow cytometry-based imaging approach for FISH in suspension (FISH-IS) was established that enables the automated analysis of several log-magnitude higher number of cells compared to the microscopy-based approaches. The rotational positioning can occur leading to discordance between spot count. As a solution of counting error from overlapping spots, in this study, a Gaussian Mixture Model based classification method is proposed. The Akaike information criterion (AIC) and Bayesian information criterion (BIC) of GMM are used as global image features of this classification method. Via Random Forest classifier, the result shows that the proposed method is able to detect closely overlapping spots which cannot be separated by existing image segmentation based spot detection methods. The experiment results show that by the proposed method we can obtain a significant improvement in spot counting accuracy.

Using methods from the data mining and machine learning literature for disease classification and prediction: A case study examining classification of heart failure sub-types

PubMed Central

Austin, Peter C.; Tu, Jack V.; Ho, Jennifer E.; Levy, Daniel; Lee, Douglas S.

2014-01-01

Objective Physicians classify patients into those with or without a specific disease. Furthermore, there is often interest in classifying patients according to disease etiology or subtype. Classification trees are frequently used to classify patients according to the presence or absence of a disease. However, classification trees can suffer from limited accuracy. In the data-mining and machine learning literature, alternate classification schemes have been developed. These include bootstrap aggregation (bagging), boosting, random forests, and support vector machines. Study design and Setting We compared the performance of these classification methods with those of conventional classification trees to classify patients with heart failure according to the following sub-types: heart failure with preserved ejection fraction (HFPEF) vs. heart failure with reduced ejection fraction (HFREF). We also compared the ability of these methods to predict the probability of the presence of HFPEF with that of conventional logistic regression. Results We found that modern, flexible tree-based methods from the data mining literature offer substantial improvement in prediction and classification of heart failure sub-type compared to conventional classification and regression trees. However, conventional logistic regression had superior performance for predicting the probability of the presence of HFPEF compared to the methods proposed in the data mining literature. Conclusion The use of tree-based methods offers superior performance over conventional classification and regression trees for predicting and classifying heart failure subtypes in a population-based sample of patients from Ontario. However, these methods do not offer substantial improvements over logistic regression for predicting the presence of HFPEF. PMID:23384592

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.

Cash for carbon: A randomized trial of payments for ecosystem services to reduce deforestation.

PubMed

Jayachandran, Seema; de Laat, Joost; Lambin, Eric F; Stanton, Charlotte Y; Audy, Robin; Thomas, Nancy E

2017-07-21

We evaluated a program of payments for ecosystem services in Uganda that offered forest-owning households annual payments of 70,000 Ugandan shillings per hectare if they conserved their forest. The program was implemented as a randomized controlled trial in 121 villages, 60 of which received the program for 2 years. The primary outcome was the change in land area covered by trees, measured by classifying high-resolution satellite imagery. We found that tree cover declined by 4.2% during the study period in treatment villages, compared to 9.1% in control villages. We found no evidence that enrollees shifted their deforestation to nearby land. We valued the delayed carbon dioxide emissions and found that this program benefit is 2.4 times as large as the program costs. Copyright © 2017 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim to original U.S. Government Works.

Accuracy assessments and areal estimates using two-phase stratified random sampling, cluster plots, and the multivariate composite estimator

Treesearch

Raymond L. Czaplewski

2000-01-01

Consider the following example of an accuracy assessment. Landsat data are used to build a thematic map of land cover for a multicounty region. The map classifier (e.g., a supervised classification algorithm) assigns each pixel into one category of land cover. The classification system includes 12 different types of forest and land cover: black spruce, balsam fir,...

Combination of lateral and PA view radiographs to study development of knee OA and associated pain

NASA Astrophysics Data System (ADS)

Minciullo, Luca; Thomson, Jessie; Cootes, Timothy F.

2017-03-01

Knee Osteoarthritis (OA) is the most common form of arthritis, affecting millions of people around the world. The effects of the disease have been studied using the shape and texture features of bones in PosteriorAnterior (PA) and Lateral radiographs separately. In this work we compare the utility of features from each view, and evaluate whether combining features from both is advantageous. We built a fully automated system to independently locate landmark points in both radiographic images using Random Forest Constrained Local Models. We extracted discriminative features from the two bony outlines using Appearance Models. The features were used to train Random Forest classifiers to solve three specific tasks: (i) OA classification, distinguishing patients with structural signs of OA from the others; (ii) predicting future onset of the disease and (iii) predicting which patients with no current pain will have a positive pain score later in a follow-up visit. Using a subset of the MOST dataset we show that the PA view has more discriminative features to classify and predict OA, while the lateral view contains features that achieve better performance in predicting pain, and that combining the features from both views gives a small improvement in accuracy of the classification compared to the individual views.

Random forest regression modelling for forest aboveground biomass estimation using RISAT-1 PolSAR and terrestrial LiDAR data

NASA Astrophysics Data System (ADS)

Mangla, Rohit; Kumar, Shashi; Nandy, Subrata

2016-05-01

SAR and LiDAR remote sensing have already shown the potential of active sensors for forest parameter retrieval. SAR sensor in its fully polarimetric mode has an advantage to retrieve scattering property of different component of forest structure and LiDAR has the capability to measure structural information with very high accuracy. This study was focused on retrieval of forest aboveground biomass (AGB) using Terrestrial Laser Scanner (TLS) based point clouds and scattering property of forest vegetation obtained from decomposition modelling of RISAT-1 fully polarimetric SAR data. TLS data was acquired for 14 plots of Timli forest range, Uttarakhand, India. The forest area is dominated by Sal trees and random sampling with plot size of 0.1 ha (31.62m*31.62m) was adopted for TLS and field data collection. RISAT-1 data was processed to retrieve SAR data based variables and TLS point clouds based 3D imaging was done to retrieve LiDAR based variables. Surface scattering, double-bounce scattering, volume scattering, helix and wire scattering were the SAR based variables retrieved from polarimetric decomposition. Tree heights and stem diameters were used as LiDAR based variables retrieved from single tree vertical height and least square circle fit methods respectively. All the variables obtained for forest plots were used as an input in a machine learning based Random Forest Regression Model, which was developed in this study for forest AGB estimation. Modelled output for forest AGB showed reliable accuracy (RMSE = 27.68 t/ha) and a good coefficient of determination (0.63) was obtained through the linear regression between modelled AGB and field-estimated AGB. The sensitivity analysis showed that the model was more sensitive for the major contributed variables (stem diameter and volume scattering) and these variables were measured from two different remote sensing techniques. This study strongly recommends the integration of SAR and LiDAR data for forest AGB estimation.

Using Landsat and a Bayesian hard classifier to study forest change in the Salmon Creek Watershed area from 1972-2013

NASA Astrophysics Data System (ADS)

Mullis, David Stone

The Salmon Creek Watershed in Sonoma County, California, USA, is home to a variety of wildlife, and many of its residents are mindful of their place in its ecology. In the past half century, several of its native and rare species have become threatened, endangered, or extinct, most notably the once common Coho salmon and Chinook salmon. The cause of this decline is believed to be a combination of global climate change, local land use, and land cover change. More specifically, the clearing of forested land to create vineyards, as well as other agricultural and residential uses, has led to a decline in biodiversity and habitat structure. I studied sub-scenes of Landsat data from 1972 to 2013 for the Salmon Creek Watershed area to estimate forest cover over this period. I used a maximum likelihood hard classifier to determine forest area, a Mahalanobis distance soft classifier to show the software's uncertainty in classification, and manually digitized forest cover to test and compare results for the 2013 30 m image. Because the earliest images were lower spatial resolution, I also tested the effects of resolution on these statistics. The images before 1985 are at 60 m spatial resolution while the later images are at 30 m resolution. Each image was processed individually and the training data were based on knowledge of the area and a mosaic of aerial photography. Each sub-scene was classified into five categories: water, forest, pasture, vineyard/orchard, and developed/barren. The research shows a decline in forest area from 1972 to around the mid-1990s, then an increase in forest area from the mid-1990s to present. The forest statistics can be helpful for conservation and restoration purposes, while the study on resolution can be helpful for landscape analysis on many levels.

An Effective Antifreeze Protein Predictor with Ensemble Classifiers and Comprehensive Sequence Descriptors.

PubMed

Yang, Runtao; Zhang, Chengjin; Gao, Rui; Zhang, Lina

2015-09-07

Antifreeze proteins (AFPs) play a pivotal role in the antifreeze effect of overwintering organisms. They have a wide range of applications in numerous fields, such as improving the production of crops and the quality of frozen foods. Accurate identification of AFPs may provide important clues to decipher the underlying mechanisms of AFPs in ice-binding and to facilitate the selection of the most appropriate AFPs for several applications. Based on an ensemble learning technique, this study proposes an AFP identification system called AFP-Ensemble. In this system, random forest classifiers are trained by different training subsets and then aggregated into a consensus classifier by majority voting. The resulting predictor yields a sensitivity of 0.892, a specificity of 0.940, an accuracy of 0.938 and a balanced accuracy of 0.916 on an independent dataset, which are far better than the results obtained by previous methods. These results reveal that AFP-Ensemble is an effective and promising predictor for large-scale determination of AFPs. The detailed feature analysis in this study may give useful insights into the molecular mechanisms of AFP-ice interactions and provide guidance for the related experimental validation. A web server has been designed to implement the proposed method.

Angiopoietin-1, Angiopoietin-2 and Bicarbonate as Diagnostic Biomarkers in Children with Severe Sepsis

PubMed Central

Wang, Kun; Bhandari, Vineet; Giuliano, John S.; O′Hern, Corey S.; Shattuck, Mark D.; Kirby, Michael

2014-01-01

Severe pediatric sepsis continues to be associated with high mortality rates in children. Thus, an important area of biomedical research is to identify biomarkers that can classify sepsis severity and outcomes. The complex and heterogeneous nature of sepsis makes the prospect of the classification of sepsis severity using a single biomarker less likely. Instead, we employ machine learning techniques to validate the use of a multiple biomarkers scoring system to determine the severity of sepsis in critically ill children. The study was based on clinical data and plasma samples provided by a tertiary care center's Pediatric Intensive Care Unit (PICU) from a group of 45 patients with varying sepsis severity at the time of admission. Canonical Correlation Analysis with the Forward Selection and Random Forests methods identified a particular set of biomarkers that included Angiopoietin-1 (Ang-1), Angiopoietin-2 (Ang-2), and Bicarbonate (HCO) as having the strongest correlations with sepsis severity. The robustness and effectiveness of these biomarkers for classifying sepsis severity were validated by constructing a linear Support Vector Machine diagnostic classifier. We also show that the concentrations of Ang-1, Ang-2, and HCO enable predictions of the time dependence of sepsis severity in children. PMID:25255212

Regulatory element-based prediction identifies new susceptibility regulatory variants for osteoporosis.

PubMed

Yao, Shi; Guo, Yan; Dong, Shan-Shan; Hao, Ruo-Han; Chen, Xiao-Feng; Chen, Yi-Xiao; Chen, Jia-Bin; Tian, Qing; Deng, Hong-Wen; Yang, Tie-Lin

2017-08-01

Despite genome-wide association studies (GWASs) have identified many susceptibility genes for osteoporosis, it still leaves a large part of missing heritability to be discovered. Integrating regulatory information and GWASs could offer new insights into the biological link between the susceptibility SNPs and osteoporosis. We generated five machine learning classifiers with osteoporosis-associated variants and regulatory features data. We gained the optimal classifier and predicted genome-wide SNPs to discover susceptibility regulatory variants. We further utilized Genetic Factors for Osteoporosis Consortium (GEFOS) and three in-house GWASs samples to validate the associations for predicted positive SNPs. The random forest classifier performed best among all machine learning methods with the F1 score of 0.8871. Using the optimized model, we predicted 37,584 candidate SNPs for osteoporosis. According to the meta-analysis results, a list of regulatory variants was significantly associated with osteoporosis after multiple testing corrections and contributed to the expression of known osteoporosis-associated protein-coding genes. In summary, combining GWASs and regulatory elements through machine learning could provide additional information for understanding the mechanism of osteoporosis. The regulatory variants we predicted will provide novel targets for etiology research and treatment of osteoporosis.

Gaps in Data and Modeling Tools for Understanding Fire and Fire Effects in Tundra Ecosystems

NASA Astrophysics Data System (ADS)

French, N. H.; Miller, M. E.; Loboda, T. V.; Jenkins, L. K.; Bourgeau-Chavez, L. L.; Suiter, A.; Hawkins, S. M.

2013-12-01

As the ecosystem science community learns more about tundra ecosystems and disturbance in tundra, a review of base data sets and ecological field data for the region shows there are many gaps that need to be filled. In this paper we will review efforts to improve our knowledge of the occurrence and impacts of fire in the North American tundra region completed under a NASA Terrestrial Ecology grant. Our main source of information is remote sensing data from satellite sensors and ecological data from past and recent field data collections by our team, collaborators, and others. Past fire occurrence is not well known for this region compared with other North American biomes. In this presentation we review an effort to use a semi-automated detection algorithm to identify past fire occurrence using the Landsat TM/ETM+ archives, pointing out some of the still-unaddressed issues for a full understanding of fire regime for the region. For this task, fires in Landsat scenes were mapped using the Random Forest classifier (Breiman 2001) to automatically detect potential burn scars. Random Forests is an ensemble classifier that employs machine learning to build a large collection of decision trees that are grown from a random selection of user supplied training data. A pixel's classification is then determined by which class receives the most 'votes' from each tree. We also review the use fire location records and existing modeling methods to quantify emissions from these fires. Based on existing maps of vegetation fuels, we used the approach developed for the Wildland Fire Emissions Information System (WFEIS; French et al. 2011) to estimate emissions across the tundra region. WFEIS employs the Consume model (http://www.fs.fed.us/pnw/fera/research/smoke/consume/index.shtml) to estimate emissions by applying empirically developed relationships between fuels, fire conditions (weather-based fire indexes), and emissions. Here again, we will review the gaps in data and modeling capability for accurate estimation of fire emissions in this region. Initial evaluation of Landsat for tundra fire characterization (Loboda et al. 2013) and successful use of the rich archive of Synthetic Aperture Radar imagery for many fire-disturbed sites in the region will be additional topics covered in this poster presentation. References: Breiman, L. 2001. Random forests. Machine Learning, 45:5-32. French, N.H.F., W.J. de Groot, L.K. Jenkins, B.. Rogers, et al. 2011. Model comparisons for estimating carbon emissions from North American wildland fire. J. Geophys. Res. 116:G00K05, doi:10.1029/2010JG001469. Loboda, T L, N H F French, C. Hight-Harf, L. Jenkins, M.E. Miller. 2013. Mapping fire extent and burn severity in Alaskan tussock tundra: An analysis of the spectral response of tundra vegetation to wildland fire. Remote Sens. Enviro. 134:194-209.

Forest cover type analysis of New England forests using innovative WorldView-2 imagery

NASA Astrophysics Data System (ADS)

Kovacs, Jenna M.

For many years, remote sensing has been used to generate land cover type maps to create a visual representation of what is occurring on the ground. One significant use of remote sensing is the identification of forest cover types. New England forests are notorious for their especially complex forest structure and as a result have been, and continue to be, a challenge when classifying forest cover types. To most accurately depict forest cover types occurring on the ground, it is essential to utilize image data that have a suitable combination of both spectral and spatial resolution. The WorldView-2 (WV2) commercial satellite, launched in 2009, is the first of its kind, having both high spectral and spatial resolutions. WV2 records eight bands of multispectral imagery, four more than the usual high spatial resolution sensors, and has a pixel size of 1.85 meters at the nadir. These additional bands have the potential to improve classification detail and classification accuracy of forest cover type maps. For this reason, WV2 imagery was utilized on its own, and in combination with Landsat 5 TM (LS5) multispectral imagery, to evaluate whether these image data could more accurately classify forest cover types. In keeping with recent developments in image analysis, an Object-Based Image Analysis (OBIA) approach was used to segment images of Pawtuckaway State Park and nearby private lands, an area representative of the typical complex forest structure found in the New England region. A Classification and Regression Tree (CART) analysis was then used to classify image segments at two levels of classification detail. Accuracies for each forest cover type map produced were generated using traditional and area-based error matrices, and additional standard accuracy measures (i.e., KAPPA) were generated. The results from this study show that there is value in analyzing imagery with both high spectral and spatial resolutions, and that WV2's new and innovative bands can be useful for the classification of complex forest structures.

Predicting general criminal recidivism in mentally disordered offenders using a random forest approach.

PubMed

Pflueger, Marlon O; Franke, Irina; Graf, Marc; Hachtel, Henning

2015-03-29

Psychiatric expert opinions are supposed to assess the accused individual's risk of reoffending based on a valid scientific foundation. In contrast to specific recidivism, general recidivism has only been poorly considered in Continental Europe; we therefore aimed to develop a valid instrument for assessing the risk of general criminal recidivism of mentally ill offenders. Data of 259 mentally ill offenders with a median time at risk of 107 months were analyzed and combined with the individuals' criminal records. We derived risk factors for general criminal recidivism and classified re-offences by using a random forest approach. In our sample of mentally ill offenders, 51% were reconvicted. The most important predictive factors for general criminal recidivism were: number of prior convictions, age, type of index offence, diversity of criminal history, and substance abuse. With our statistical approach we were able to correctly identify 58-95% of all reoffenders and 65-97% of all committed offences (AUC = .90). Our study presents a new statistical approach to forensic-psychiatric risk-assessment, allowing experts to evaluate general risk of reoffending in mentally disordered individuals, with a special focus on high-risk groups. This approach might serve not only for expert opinions in court, but also for risk management strategies and therapeutic interventions.

Predicting human liver microsomal stability with machine learning techniques.

PubMed

Sakiyama, Yojiro; Yuki, Hitomi; Moriya, Takashi; Hattori, Kazunari; Suzuki, Misaki; Shimada, Kaoru; Honma, Teruki

2008-02-01

To ensure a continuing pipeline in pharmaceutical research, lead candidates must possess appropriate metabolic stability in the drug discovery process. In vitro ADMET (absorption, distribution, metabolism, elimination, and toxicity) screening provides us with useful information regarding the metabolic stability of compounds. However, before the synthesis stage, an efficient process is required in order to deal with the vast quantity of data from large compound libraries and high-throughput screening. Here we have derived a relationship between the chemical structure and its metabolic stability for a data set of in-house compounds by means of various in silico machine learning such as random forest, support vector machine (SVM), logistic regression, and recursive partitioning. For model building, 1952 proprietary compounds comprising two classes (stable/unstable) were used with 193 descriptors calculated by Molecular Operating Environment. The results using test compounds have demonstrated that all classifiers yielded satisfactory results (accuracy > 0.8, sensitivity > 0.9, specificity > 0.6, and precision > 0.8). Above all, classification by random forest as well as SVM yielded kappa values of approximately 0.7 in an independent validation set, slightly higher than other classification tools. These results suggest that nonlinear/ensemble-based classification methods might prove useful in the area of in silico ADME modeling.

An evaluation of supervised classifiers for indirectly detecting salt-affected areas at irrigation scheme level

NASA Astrophysics Data System (ADS)

Muller, Sybrand Jacobus; van Niekerk, Adriaan

2016-07-01

Soil salinity often leads to reduced crop yield and quality and can render soils barren. Irrigated areas are particularly at risk due to intensive cultivation and secondary salinization caused by waterlogging. Regular monitoring of salt accumulation in irrigation schemes is needed to keep its negative effects under control. The dynamic spatial and temporal characteristics of remote sensing can provide a cost-effective solution for monitoring salt accumulation at irrigation scheme level. This study evaluated a range of pan-fused SPOT-5 derived features (spectral bands, vegetation indices, image textures and image transformations) for classifying salt-affected areas in two distinctly different irrigation schemes in South Africa, namely Vaalharts and Breede River. The relationship between the input features and electro conductivity measurements were investigated using regression modelling (stepwise linear regression, partial least squares regression, curve fit regression modelling) and supervised classification (maximum likelihood, nearest neighbour, decision tree analysis, support vector machine and random forests). Classification and regression trees and random forest were used to select the most important features for differentiating salt-affected and unaffected areas. The results showed that the regression analyses produced weak models (<0.4 R squared). Better results were achieved using the supervised classifiers, but the algorithms tend to over-estimate salt-affected areas. A key finding was that none of the feature sets or classification algorithms stood out as being superior for monitoring salt accumulation at irrigation scheme level. This was attributed to the large variations in the spectral responses of different crops types at different growing stages, coupled with their individual tolerances to saline conditions.

A Model-Based Approach to Inventory Stratification

Treesearch

Ronald E. McRoberts

2006-01-01

Forest inventory programs report estimates of forest variables for areas of interest ranging in size from municipalities to counties to States and Provinces. Classified satellite imagery has been shown to be an effective source of ancillary data that, when used with stratified estimation techniques, contributes to increased precision with little corresponding increase...

National forest trail users: planning for recreation opportunities

Treesearch

John J. Daigle; Alan E. Watson; Glenn E. Haas

1994-01-01

National forest trail users in four geographical regions of the United States are described based on participation in clusters of recreation activities. Visitors are classified into day hiking, undeveloped recreation, and two developed camping and hiking activity clusters for the Appalachian, Pacific, Rocky Mountain, and Southwestern regions. Distance and time traveled...

A hybrid network-based method for the detection of disease-related genes

NASA Astrophysics Data System (ADS)

Cui, Ying; Cai, Meng; Dai, Yang; Stanley, H. Eugene

2018-02-01

Detecting disease-related genes is crucial in disease diagnosis and drug design. The accepted view is that neighbors of a disease-causing gene in a molecular network tend to cause the same or similar diseases, and network-based methods have been recently developed to identify novel hereditary disease-genes in available biomedical networks. Despite the steady increase in the discovery of disease-associated genes, there is still a large fraction of disease genes that remains under the tip of the iceberg. In this paper we exploit the topological properties of the protein-protein interaction (PPI) network to detect disease-related genes. We compute, analyze, and compare the topological properties of disease genes with non-disease genes in PPI networks. We also design an improved random forest classifier based on these network topological features, and a cross-validation test confirms that our method performs better than previous similar studies.

Can single classifiers be as useful as model ensembles to produce benthic seabed substratum maps?

NASA Astrophysics Data System (ADS)

Turner, Joseph A.; Babcock, Russell C.; Hovey, Renae; Kendrick, Gary A.

2018-05-01

Numerous machine-learning classifiers are available for benthic habitat map production, which can lead to different results. This study highlights the performance of the Random Forest (RF) classifier, which was significantly better than Classification Trees (CT), Naïve Bayes (NB), and a multi-model ensemble in terms of overall accuracy, Balanced Error Rate (BER), Kappa, and area under the curve (AUC) values. RF accuracy was often higher than 90% for each substratum class, even at the most detailed level of the substratum classification and AUC values also indicated excellent performance (0.8-1). Total agreement between classifiers was high at the broadest level of classification (75-80%) when differentiating between hard and soft substratum. However, this sharply declined as the number of substratum categories increased (19-45%) including a mix of rock, gravel, pebbles, and sand. The model ensemble, produced from the results of all three classifiers by majority voting, did not show any increase in predictive performance when compared to the single RF classifier. This study shows how a single classifier may be sufficient to produce benthic seabed maps and model ensembles of multiple classifiers.

Forest Resource Information System (FRIS)

NASA Technical Reports Server (NTRS)

1983-01-01

The technological and economical feasibility of using multispectral digital image data as acquired from the LANDSAT satellites in an ongoing operational forest information system was evaluated. Computer compatible multispectral scanner data secured from the LANDSAT satellites were demonstrated to be a significant contributor to ongoing information systems by providing the added dimensions of synoptic and repeat coverage of the Earth's surface. Major forest cover types of conifer, deciduous, mixed conifer-deciduous and non-forest, were classified well within the bounds of the statistical accuracy of the ground sample. Further, when overlayed with existing maps, the acreage of cover type retains a high level of positional integrity. Maps were digitized by a graphics design system, overlayed and registered onto LANDSAT imagery such that the map data with associated attributes were displayed on the image. Once classified, the analysis results were converted back to map form as a cover type of information. Existing tabular information as represented by inventory is registered geographically to the map base through a vendor provided data management system. The notion of a geographical reference base (map) providing the framework to which imagery and tabular data bases are registered and where each of the three functions of imagery, maps and inventory can be accessed singly or in combination is the very essence of the forest resource information system design.

Multi-fractal detrended texture feature for brain tumor classification

NASA Astrophysics Data System (ADS)

Reza, Syed M. S.; Mays, Randall; Iftekharuddin, Khan M.

2015-03-01

We propose a novel non-invasive brain tumor type classification using Multi-fractal Detrended Fluctuation Analysis (MFDFA) [1] in structural magnetic resonance (MR) images. This preliminary work investigates the efficacy of the MFDFA features along with our novel texture feature known as multifractional Brownian motion (mBm) [2] in classifying (grading) brain tumors as High Grade (HG) and Low Grade (LG). Based on prior performance, Random Forest (RF) [3] is employed for tumor grading using two different datasets such as BRATS-2013 [4] and BRATS-2014 [5]. Quantitative scores such as precision, recall, accuracy are obtained using the confusion matrix. On an average 90% precision and 85% recall from the inter-dataset cross-validation confirm the efficacy of the proposed method.

Sequence Based Prediction of Antioxidant Proteins Using a Classifier Selection Strategy

PubMed Central

Zhang, Lina; Zhang, Chengjin; Gao, Rui; Yang, Runtao; Song, Qing

2016-01-01

Antioxidant proteins perform significant functions in maintaining oxidation/antioxidation balance and have potential therapies for some diseases. Accurate identification of antioxidant proteins could contribute to revealing physiological processes of oxidation/antioxidation balance and developing novel antioxidation-based drugs. In this study, an ensemble method is presented to predict antioxidant proteins with hybrid features, incorporating SSI (Secondary Structure Information), PSSM (Position Specific Scoring Matrix), RSA (Relative Solvent Accessibility), and CTD (Composition, Transition, Distribution). The prediction results of the ensemble predictor are determined by an average of prediction results of multiple base classifiers. Based on a classifier selection strategy, we obtain an optimal ensemble classifier composed of RF (Random Forest), SMO (Sequential Minimal Optimization), NNA (Nearest Neighbor Algorithm), and J48 with an accuracy of 0.925. A Relief combined with IFS (Incremental Feature Selection) method is adopted to obtain optimal features from hybrid features. With the optimal features, the ensemble method achieves improved performance with a sensitivity of 0.95, a specificity of 0.93, an accuracy of 0.94, and an MCC (Matthew’s Correlation Coefficient) of 0.880, far better than the existing method. To evaluate the prediction performance objectively, the proposed method is compared with existing methods on the same independent testing dataset. Encouragingly, our method performs better than previous studies. In addition, our method achieves more balanced performance with a sensitivity of 0.878 and a specificity of 0.860. These results suggest that the proposed ensemble method can be a potential candidate for antioxidant protein prediction. For public access, we develop a user-friendly web server for antioxidant protein identification that is freely accessible at http://antioxidant.weka.cc. PMID:27662651

On the classification techniques in data mining for microarray data classification

NASA Astrophysics Data System (ADS)

Aydadenta, Husna; Adiwijaya

2018-03-01

Cancer is one of the deadly diseases, according to data from WHO by 2015 there are 8.8 million more deaths caused by cancer, and this will increase every year if not resolved earlier. Microarray data has become one of the most popular cancer-identification studies in the field of health, since microarray data can be used to look at levels of gene expression in certain cell samples that serve to analyze thousands of genes simultaneously. By using data mining technique, we can classify the sample of microarray data thus it can be identified with cancer or not. In this paper we will discuss some research using some data mining techniques using microarray data, such as Support Vector Machine (SVM), Artificial Neural Network (ANN), Naive Bayes, k-Nearest Neighbor (kNN), and C4.5, and simulation of Random Forest algorithm with technique of reduction dimension using Relief. The result of this paper show performance measure (accuracy) from classification algorithm (SVM, ANN, Naive Bayes, kNN, C4.5, and Random Forets).The results in this paper show the accuracy of Random Forest algorithm higher than other classification algorithms (Support Vector Machine (SVM), Artificial Neural Network (ANN), Naive Bayes, k-Nearest Neighbor (kNN), and C4.5). It is hoped that this paper can provide some information about the speed, accuracy, performance and computational cost generated from each Data Mining Classification Technique based on microarray data.

Applying genetic algorithms to set the optimal combination of forest fire related variables and model forest fire susceptibility based on data mining models. The case of Dayu County, China.

PubMed

Hong, Haoyuan; Tsangaratos, Paraskevas; Ilia, Ioanna; Liu, Junzhi; Zhu, A-Xing; Xu, Chong

2018-07-15

The main objective of the present study was to utilize Genetic Algorithms (GA) in order to obtain the optimal combination of forest fire related variables and apply data mining methods for constructing a forest fire susceptibility map. In the proposed approach, a Random Forest (RF) and a Support Vector Machine (SVM) was used to produce a forest fire susceptibility map for the Dayu County which is located in southwest of Jiangxi Province, China. For this purpose, historic forest fires and thirteen forest fire related variables were analyzed, namely: elevation, slope angle, aspect, curvature, land use, soil cover, heat load index, normalized difference vegetation index, mean annual temperature, mean annual wind speed, mean annual rainfall, distance to river network and distance to road network. The Natural Break and the Certainty Factor method were used to classify and weight the thirteen variables, while a multicollinearity analysis was performed to determine the correlation among the variables and decide about their usability. The optimal set of variables, determined by the GA limited the number of variables into eight excluding from the analysis, aspect, land use, heat load index, distance to river network and mean annual rainfall. The performance of the forest fire models was evaluated by using the area under the Receiver Operating Characteristic curve (ROC-AUC) based on the validation dataset. Overall, the RF models gave higher AUC values. Also the results showed that the proposed optimized models outperform the original models. Specifically, the optimized RF model gave the best results (0.8495), followed by the original RF (0.8169), while the optimized SVM gave lower values (0.7456) than the RF, however higher than the original SVM (0.7148) model. The study highlights the significance of feature selection techniques in forest fire susceptibility, whereas data mining methods could be considered as a valid approach for forest fire susceptibility modeling. Copyright © 2018 Elsevier B.V. All rights reserved.

Classification of large-sized hyperspectral imagery using fast machine learning algorithms

NASA Astrophysics Data System (ADS)

Xia, Junshi; Yokoya, Naoto; Iwasaki, Akira

2017-07-01

We present a framework of fast machine learning algorithms in the context of large-sized hyperspectral images classification from the theoretical to a practical viewpoint. In particular, we assess the performance of random forest (RF), rotation forest (RoF), and extreme learning machine (ELM) and the ensembles of RF and ELM. These classifiers are applied to two large-sized hyperspectral images and compared to the support vector machines. To give the quantitative analysis, we pay attention to comparing these methods when working with high input dimensions and a limited/sufficient training set. Moreover, other important issues such as the computational cost and robustness against the noise are also discussed.

Prediction of Chemical Function: Model Development and ...

EPA Pesticide Factsheets

The United States Environmental Protection Agency’s Exposure Forecaster (ExpoCast) project is developing both statistical and mechanism-based computational models for predicting exposures to thousands of chemicals, including those in consumer products. The high-throughput (HT) screening-level exposures developed under ExpoCast can be combined with HT screening (HTS) bioactivity data for the risk-based prioritization of chemicals for further evaluation. The functional role (e.g. solvent, plasticizer, fragrance) that a chemical performs can drive both the types of products in which it is found and the concentration in which it is present and therefore impacting exposure potential. However, critical chemical use information (including functional role) is lacking for the majority of commercial chemicals for which exposure estimates are needed. A suite of machine-learning based models for classifying chemicals in terms of their likely functional roles in products based on structure were developed. This effort required collection, curation, and harmonization of publically-available data sources of chemical functional use information from government and industry bodies. Physicochemical and structure descriptor data were generated for chemicals with function data. Machine-learning classifier models for function were then built in a cross-validated manner from the descriptor/function data using the method of random forests. The models were applied to: 1) predict chemi

Producing landslide susceptibility maps by utilizing machine learning methods. The case of Finikas catchment basin, North Peloponnese, Greece.

NASA Astrophysics Data System (ADS)

Tsangaratos, Paraskevas; Ilia, Ioanna; Loupasakis, Constantinos; Papadakis, Michalis; Karimalis, Antonios

2017-04-01

The main objective of the present study was to apply two machine learning methods for the production of a landslide susceptibility map in the Finikas catchment basin, located in North Peloponnese, Greece and to compare their results. Specifically, Logistic Regression and Random Forest were utilized, based on a database of 40 sites classified into two categories, non-landslide and landslide areas that were separated into a training dataset (70% of the total data) and a validation dataset (remaining 30%). The identification of the areas was established by analyzing airborne imagery, extensive field investigation and the examination of previous research studies. Six landslide related variables were analyzed, namely: lithology, elevation, slope, aspect, distance to rivers and distance to faults. Within the Finikas catchment basin most of the reported landslides were located along the road network and within the residential complexes, classified as rotational and translational slides, and rockfalls, mainly caused due to the physical conditions and the general geotechnical behavior of the geological formation that cover the area. Each landslide susceptibility map was reclassified by applying the Geometric Interval classification technique into five classes, namely: very low susceptibility, low susceptibility, moderate susceptibility, high susceptibility, and very high susceptibility. The comparison and validation of the outcomes of each model were achieved using statistical evaluation measures, the receiving operating characteristic and the area under the success and predictive rate curves. The computation process was carried out using RStudio an integrated development environment for R language and ArcGIS 10.1 for compiling the data and producing the landslide susceptibility maps. From the outcomes of the Logistic Regression analysis it was induced that the highest b coefficient is allocated to lithology and slope, which was 2.8423 and 1.5841, respectively. From the estimation of the mean decrease in Gini coefficient performed during the application of Random Forest and the mean decrease in accuracy the most important variable is slope followed by lithology, aspect, elevation, distance from river network, and distance from faults, while the most used variables during the training phase were the variable aspect (21.45%), slope (20.53%) and lithology (19.84%). The outcomes of the analysis are consistent with previous studies concerning the area of research, which have indicated the high influence of lithology and slope in the manifestation of landslides. High percentage of landslide occurrence has been observed in Plio-Pleistocene sediments, flysch formations, and Cretaceous limestone. Also the presences of landslides have been associated with the degree of weathering and fragmentation, the orientation of the discontinuities surfaces and the intense morphological relief. The most accurate model was Random Forest which identified correctly 92.00% of the instances during the training phase, followed by the Logistic Regression 89.00%. The same pattern of accuracy was calculated during the validation phase, in which the Random Forest achieved a classification accuracy of 93.00%, while the Logistic Regression model achieved an accuracy of 91.00%. In conclusion, the outcomes of the study could be a useful cartographic product to local authorities and government agencies during the implementation of successful decision-making and land use planning strategies. Keywords: Landslide Susceptibility, Logistic Regression, Random Forest, GIS, Greece.

Assessment of vegetation change in a fire-altered forest landscape

NASA Technical Reports Server (NTRS)

Jakubauskas, Mark E.; Lulla, Kamlesh P.; Mausel, Paul W.

1990-01-01

This research focused on determining the degree to which differences in burn severity relate to postfire vegetative cover within a Michigan pine forest. Landsat MSS data from June 1973 and TM data from October 1982 were classified using an unsupervised approach to create prefire and postfire cover maps of the study area. Using a raster-based geographic information system (GIS), the maps were compared, and a map of vegetation change was created. An IR/red band ratio from a June 1980 Landsat scene was classified to create a map of three degres of burn severity, which was then compared with the vegetation change map using a GIS. Classification comparisons of pine and deciduous forest classes (1973 to 1982) revealed that the most change in vegetation occurred in areas subjected to the most intense burn. Two classes of regenerating forest comprised the majority of the change, while the remaining change was associated with shrub vegetation or another forest class.

Potential reduction in terrestrial salamander ranges associated with Marcellus shale development

USGS Publications Warehouse

Brand, Adrianne B,; Wiewel, Amber N. M.; Grant, Evan H. Campbell

2014-01-01

Natural gas production from the Marcellus shale is rapidly increasing in the northeastern United States. Most of the endemic terrestrial salamander species in the region are classified as ‘globally secure’ by the IUCN, primarily because much of their ranges include state- and federally protected lands, which have been presumed to be free from habitat loss. However, the proposed and ongoing development of the Marcellus gas resources may result in significant range restrictions for these and other terrestrial forest salamanders. To begin to address the gaps in our knowledge of the direct impacts of shale gas development, we developed occurrence models for five species of terrestrial plethodontid salamanders found largely within the Marcellus shale play. We predicted future Marcellus shale development under several scenarios. Under scenarios of 10,000, 20,000, and 50,000 new gas wells, we predict 4%, 8%, and 20% forest loss, respectively, within the play. Predictions of habitat loss vary among species, but in general, Plethodon electromorphus and Plethodonwehrlei are predicted to lose the greatest proportion of forested habitat within their ranges if future Marcellus development is based on characteristics of the shale play. If development is based on current well locations,Plethodonrichmondi is predicted to lose the greatest proportion of habitat. Models showed high uncertainty in species’ ranges and emphasize the need for distribution data collected by widespread and repeated, randomized surveys.

CW-SSIM kernel based random forest for image classification

NASA Astrophysics Data System (ADS)

Fan, Guangzhe; Wang, Zhou; Wang, Jiheng

2010-07-01

Complex wavelet structural similarity (CW-SSIM) index has been proposed as a powerful image similarity metric that is robust to translation, scaling and rotation of images, but how to employ it in image classification applications has not been deeply investigated. In this paper, we incorporate CW-SSIM as a kernel function into a random forest learning algorithm. This leads to a novel image classification approach that does not require a feature extraction or dimension reduction stage at the front end. We use hand-written digit recognition as an example to demonstrate our algorithm. We compare the performance of the proposed approach with random forest learning based on other kernels, including the widely adopted Gaussian and the inner product kernels. Empirical evidences show that the proposed method is superior in its classification power. We also compared our proposed approach with the direct random forest method without kernel and the popular kernel-learning method support vector machine. Our test results based on both simulated and realworld data suggest that the proposed approach works superior to traditional methods without the feature selection procedure.

Beyond the hype: deep neural networks outperform established methods using a ChEMBL bioactivity benchmark set.

PubMed

Lenselink, Eelke B; Ten Dijke, Niels; Bongers, Brandon; Papadatos, George; van Vlijmen, Herman W T; Kowalczyk, Wojtek; IJzerman, Adriaan P; van Westen, Gerard J P

2017-08-14

The increase of publicly available bioactivity data in recent years has fueled and catalyzed research in chemogenomics, data mining, and modeling approaches. As a direct result, over the past few years a multitude of different methods have been reported and evaluated, such as target fishing, nearest neighbor similarity-based methods, and Quantitative Structure Activity Relationship (QSAR)-based protocols. However, such studies are typically conducted on different datasets, using different validation strategies, and different metrics. In this study, different methods were compared using one single standardized dataset obtained from ChEMBL, which is made available to the public, using standardized metrics (BEDROC and Matthews Correlation Coefficient). Specifically, the performance of Naïve Bayes, Random Forests, Support Vector Machines, Logistic Regression, and Deep Neural Networks was assessed using QSAR and proteochemometric (PCM) methods. All methods were validated using both a random split validation and a temporal validation, with the latter being a more realistic benchmark of expected prospective execution. Deep Neural Networks are the top performing classifiers, highlighting the added value of Deep Neural Networks over other more conventional methods. Moreover, the best method ('DNN_PCM') performed significantly better at almost one standard deviation higher than the mean performance. Furthermore, Multi-task and PCM implementations were shown to improve performance over single task Deep Neural Networks. Conversely, target prediction performed almost two standard deviations under the mean performance. Random Forests, Support Vector Machines, and Logistic Regression performed around mean performance. Finally, using an ensemble of DNNs, alongside additional tuning, enhanced the relative performance by another 27% (compared with unoptimized 'DNN_PCM'). Here, a standardized set to test and evaluate different machine learning algorithms in the context of multi-task learning is offered by providing the data and the protocols. Graphical Abstract .

DNABP: Identification of DNA-Binding Proteins Based on Feature Selection Using a Random Forest and Predicting Binding Residues.

PubMed

Ma, Xin; Guo, Jing; Sun, Xiao

2016-01-01

DNA-binding proteins are fundamentally important in cellular processes. Several computational-based methods have been developed to improve the prediction of DNA-binding proteins in previous years. However, insufficient work has been done on the prediction of DNA-binding proteins from protein sequence information. In this paper, a novel predictor, DNABP (DNA-binding proteins), was designed to predict DNA-binding proteins using the random forest (RF) classifier with a hybrid feature. The hybrid feature contains two types of novel sequence features, which reflect information about the conservation of physicochemical properties of the amino acids, and the binding propensity of DNA-binding residues and non-binding propensities of non-binding residues. The comparisons with each feature demonstrated that these two novel features contributed most to the improvement in predictive ability. Furthermore, to improve the prediction performance of the DNABP model, feature selection using the minimum redundancy maximum relevance (mRMR) method combined with incremental feature selection (IFS) was carried out during the model construction. The results showed that the DNABP model could achieve 86.90% accuracy, 83.76% sensitivity, 90.03% specificity and a Matthews correlation coefficient of 0.727. High prediction accuracy and performance comparisons with previous research suggested that DNABP could be a useful approach to identify DNA-binding proteins from sequence information. The DNABP web server system is freely available at http://www.cbi.seu.edu.cn/DNABP/.

A comparison of machine learning techniques for survival prediction in breast cancer

PubMed Central

2011-01-01

Background The ability to accurately classify cancer patients into risk classes, i.e. to predict the outcome of the pathology on an individual basis, is a key ingredient in making therapeutic decisions. In recent years gene expression data have been successfully used to complement the clinical and histological criteria traditionally used in such prediction. Many "gene expression signatures" have been developed, i.e. sets of genes whose expression values in a tumor can be used to predict the outcome of the pathology. Here we investigate the use of several machine learning techniques to classify breast cancer patients using one of such signatures, the well established 70-gene signature. Results We show that Genetic Programming performs significantly better than Support Vector Machines, Multilayered Perceptrons and Random Forests in classifying patients from the NKI breast cancer dataset, and comparably to the scoring-based method originally proposed by the authors of the 70-gene signature. Furthermore, Genetic Programming is able to perform an automatic feature selection. Conclusions Since the performance of Genetic Programming is likely to be improvable compared to the out-of-the-box approach used here, and given the biological insight potentially provided by the Genetic Programming solutions, we conclude that Genetic Programming methods are worth further investigation as a tool for cancer patient classification based on gene expression data. PMID:21569330

Comparative evaluation of support vector machine classification for computer aided detection of breast masses in mammography

NASA Astrophysics Data System (ADS)

Lesniak, J. M.; Hupse, R.; Blanc, R.; Karssemeijer, N.; Székely, G.

2012-08-01

False positive (FP) marks represent an obstacle for effective use of computer-aided detection (CADe) of breast masses in mammography. Typically, the problem can be approached either by developing more discriminative features or by employing different classifier designs. In this paper, the usage of support vector machine (SVM) classification for FP reduction in CADe is investigated, presenting a systematic quantitative evaluation against neural networks, k-nearest neighbor classification, linear discriminant analysis and random forests. A large database of 2516 film mammography examinations and 73 input features was used to train the classifiers and evaluate for their performance on correctly diagnosed exams as well as false negatives. Further, classifier robustness was investigated using varying training data and feature sets as input. The evaluation was based on the mean exam sensitivity in 0.05-1 FPs on normals on the free-response receiver operating characteristic curve (FROC), incorporated into a tenfold cross validation framework. It was found that SVM classification using a Gaussian kernel offered significantly increased detection performance (P = 0.0002) compared to the reference methods. Varying training data and input features, SVMs showed improved exploitation of large feature sets. It is concluded that with the SVM-based CADe a significant reduction of FPs is possible outperforming other state-of-the-art approaches for breast mass CADe.

Sequence Based Prediction of DNA-Binding Proteins Based on Hybrid Feature Selection Using Random Forest and Gaussian Naïve Bayes

PubMed Central

Lou, Wangchao; Wang, Xiaoqing; Chen, Fan; Chen, Yixiao; Jiang, Bo; Zhang, Hua

2014-01-01

Developing an efficient method for determination of the DNA-binding proteins, due to their vital roles in gene regulation, is becoming highly desired since it would be invaluable to advance our understanding of protein functions. In this study, we proposed a new method for the prediction of the DNA-binding proteins, by performing the feature rank using random forest and the wrapper-based feature selection using forward best-first search strategy. The features comprise information from primary sequence, predicted secondary structure, predicted relative solvent accessibility, and position specific scoring matrix. The proposed method, called DBPPred, used Gaussian naïve Bayes as the underlying classifier since it outperformed five other classifiers, including decision tree, logistic regression, k-nearest neighbor, support vector machine with polynomial kernel, and support vector machine with radial basis function. As a result, the proposed DBPPred yields the highest average accuracy of 0.791 and average MCC of 0.583 according to the five-fold cross validation with ten runs on the training benchmark dataset PDB594. Subsequently, blind tests on the independent dataset PDB186 by the proposed model trained on the entire PDB594 dataset and by other five existing methods (including iDNA-Prot, DNA-Prot, DNAbinder, DNABIND and DBD-Threader) were performed, resulting in that the proposed DBPPred yielded the highest accuracy of 0.769, MCC of 0.538, and AUC of 0.790. The independent tests performed by the proposed DBPPred on completely a large non-DNA binding protein dataset and two RNA binding protein datasets also showed improved or comparable quality when compared with the relevant prediction methods. Moreover, we observed that majority of the selected features by the proposed method are statistically significantly different between the mean feature values of the DNA-binding and the non DNA-binding proteins. All of the experimental results indicate that the proposed DBPPred can be an alternative perspective predictor for large-scale determination of DNA-binding proteins. PMID:24475169

Drivers of community assembly in tropical forest restoration sites: role of local environment, landscape, and space.

PubMed

Audino, Lívia D; Murphy, Stephen J; Zambaldi, Ludimila; Louzada, Julio; Comita, Liza S

2017-09-01

There is increasing recognition that community assembly theory can offer valuable insights for ecological restoration. We studied community assembly processes following tropical forest restoration efforts, using dung beetles (Scarabaeinae) as a focal taxon to investigate taxonomic and functional patterns of biodiversity recovery. We evaluated the relative importance of the local environment (i.e., canopy cover, understory cover, tree basal area, and soil texture), landscape context (i.e., habitat patch proximity and availability and percentage of surrounding area classified as natural forest or Eucalyptus spp. plantation), and space (i.e., spatial proximity of the study areas to estimate dispersal limitation or unmeasured spatially structured processes) on dung beetle species and functional trait composition across a gradient of 15 restoration sites in Brazilian Atlantic Forest. We also assessed which factors were the primary determinants in the establishment of individual dung beetle functional groups, classified according to size, food relocation habit, diet, and period of flight activity. Both species and functional trait composition were most strongly influenced by the local environment, indicating that assembly was predominantly driven by niche-based processes. Most of the variation explained by space was co-explained by local environment and landscape context, ruling out a strong influence of dispersal limitation and random colonization on assembly following restoration. In addition, nearly all of the variance explained by landscape context was co-explained by local environment, suggesting that arrival and establishment at a site depends on both local and landscape-scale environmental factors. Despite strong evidence for niche-based assembly, a large amount of variation remained unexplained in all models, suggesting that stochastic processes and/or unmeasured environmental variables also play an important role. The relative importance of local environment, landscape context, and space changed considerably when analyzing the assembly mechanisms of each functional group separately. Therefore, to recover distinct functional traits in restoration sites, it may be necessary to manipulate different components of the local environment and surrounding landscape. Overall, this study shows that assembly rules can help to better understand recovery processes, enabling improvement of future restoration efforts. © 2017 by the Ecological Society of America.

Supervised Detection of Anomalous Light Curves in Massive Astronomical Catalogs

NASA Astrophysics Data System (ADS)

Nun, Isadora; Pichara, Karim; Protopapas, Pavlos; Kim, Dae-Won

2014-09-01

The development of synoptic sky surveys has led to a massive amount of data for which resources needed for analysis are beyond human capabilities. In order to process this information and to extract all possible knowledge, machine learning techniques become necessary. Here we present a new methodology to automatically discover unknown variable objects in large astronomical catalogs. With the aim of taking full advantage of all information we have about known objects, our method is based on a supervised algorithm. In particular, we train a random forest classifier using known variability classes of objects and obtain votes for each of the objects in the training set. We then model this voting distribution with a Bayesian network and obtain the joint voting distribution among the training objects. Consequently, an unknown object is considered as an outlier insofar it has a low joint probability. By leaving out one of the classes on the training set, we perform a validity test and show that when the random forest classifier attempts to classify unknown light curves (the class left out), it votes with an unusual distribution among the classes. This rare voting is detected by the Bayesian network and expressed as a low joint probability. Our method is suitable for exploring massive data sets given that the training process is performed offline. We tested our algorithm on 20 million light curves from the MACHO catalog and generated a list of anomalous candidates. After analysis, we divided the candidates into two main classes of outliers: artifacts and intrinsic outliers. Artifacts were principally due to air mass variation, seasonal variation, bad calibration, or instrumental errors and were consequently removed from our outlier list and added to the training set. After retraining, we selected about 4000 objects, which we passed to a post-analysis stage by performing a cross-match with all publicly available catalogs. Within these candidates we identified certain known but rare objects such as eclipsing Cepheids, blue variables, cataclysmic variables, and X-ray sources. For some outliers there was no additional information. Among them we identified three unknown variability types and a few individual outliers that will be followed up in order to perform a deeper analysis.

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

Nun, Isadora; Pichara, Karim; Protopapas, Pavlos

The development of synoptic sky surveys has led to a massive amount of data for which resources needed for analysis are beyond human capabilities. In order to process this information and to extract all possible knowledge, machine learning techniques become necessary. Here we present a new methodology to automatically discover unknown variable objects in large astronomical catalogs. With the aim of taking full advantage of all information we have about known objects, our method is based on a supervised algorithm. In particular, we train a random forest classifier using known variability classes of objects and obtain votes for each ofmore » the objects in the training set. We then model this voting distribution with a Bayesian network and obtain the joint voting distribution among the training objects. Consequently, an unknown object is considered as an outlier insofar it has a low joint probability. By leaving out one of the classes on the training set, we perform a validity test and show that when the random forest classifier attempts to classify unknown light curves (the class left out), it votes with an unusual distribution among the classes. This rare voting is detected by the Bayesian network and expressed as a low joint probability. Our method is suitable for exploring massive data sets given that the training process is performed offline. We tested our algorithm on 20 million light curves from the MACHO catalog and generated a list of anomalous candidates. After analysis, we divided the candidates into two main classes of outliers: artifacts and intrinsic outliers. Artifacts were principally due to air mass variation, seasonal variation, bad calibration, or instrumental errors and were consequently removed from our outlier list and added to the training set. After retraining, we selected about 4000 objects, which we passed to a post-analysis stage by performing a cross-match with all publicly available catalogs. Within these candidates we identified certain known but rare objects such as eclipsing Cepheids, blue variables, cataclysmic variables, and X-ray sources. For some outliers there was no additional information. Among them we identified three unknown variability types and a few individual outliers that will be followed up in order to perform a deeper analysis.« less

Forest Products Laboratory natural finish

Treesearch

J. M. Black; D. F. Laughnan; E. A. Mraz

1979-01-01

A simple and durable exterior natural finish developed at the Forest Products Laboratory is described. The finish is classified as a semi-transparent oil-base penetrating stain that effectively retains much of the natural grain and texture of wood when exposed to the weather. The directions for preparation are included as are the recommendations for application to both...

The use of space and high altitude aerial photography to classify forest land and to detect forest disturbances

NASA Technical Reports Server (NTRS)

Aldrich, R. C.; Greentree, W. J.; Heller, R. C.; Norick, N. X.

1970-01-01

In October 1969, an investigation was begun near Atlanta, Georgia, to explore the possibilities of developing predictors for forest land and stand condition classifications using space photography. It has been found that forest area can be predicted with reasonable accuracy on space photographs using ocular techniques. Infrared color film is the best single multiband sensor for this purpose. Using the Apollo 9 infrared color photographs taken in March 1969 photointerpreters were able to predict forest area for small units consistently within 5 to 10 percent of ground truth. Approximately 5,000 density data points were recorded for 14 scan lines selected at random from five study blocks. The mean densities and standard deviations were computed for 13 separate land use classes. The results indicate that forest area cannot be separated from other land uses with a high degree of accuracy using optical film density alone. If, however, densities derived by introducing red, green, and blue cutoff filters in the optical system of the microdensitometer are combined with their differences and their ratios in regression analysis techniques, there is a good possibility of discriminating forest from all other classes.

Remote sensing change detection methods to track deforestation and growth in threatened rainforests in Madre de Dios, Peru

USGS Publications Warehouse

Shermeyer, Jacob S.; Haack, Barry N.

2015-01-01

Two forestry-change detection methods are described, compared, and contrasted for estimating deforestation and growth in threatened forests in southern Peru from 2000 to 2010. The methods used in this study rely on freely available data, including atmospherically corrected Landsat 5 Thematic Mapper and Moderate Resolution Imaging Spectroradiometer (MODIS) vegetation continuous fields (VCF). The two methods include a conventional supervised signature extraction method and a unique self-calibrating method called MODIS VCF guided forest/nonforest (FNF) masking. The process chain for each of these methods includes a threshold classification of MODIS VCF, training data or signature extraction, signature evaluation, k-nearest neighbor classification, analyst-guided reclassification, and postclassification image differencing to generate forest change maps. Comparisons of all methods were based on an accuracy assessment using 500 validation pixels. Results of this accuracy assessment indicate that FNF masking had a 5% higher overall accuracy and was superior to conventional supervised classification when estimating forest change. Both methods succeeded in classifying persistently forested and nonforested areas, and both had limitations when classifying forest change.

Automatic segmentation of lumbar vertebrae in CT images

NASA Astrophysics Data System (ADS)

Kulkarni, Amruta; Raina, Akshita; Sharifi Sarabi, Mona; Ahn, Christine S.; Babayan, Diana; Gaonkar, Bilwaj; Macyszyn, Luke; Raghavendra, Cauligi

2017-03-01

Lower back pain is one of the most prevalent disorders in the developed/developing world. However, its etiology is poorly understood and treatment is often determined subjectively. In order to quantitatively study the emergence and evolution of back pain, it is necessary to develop consistently measurable markers for pathology. Imaging based measures offer one solution to this problem. The development of imaging based on quantitative biomarkers for the lower back necessitates automated techniques to acquire this data. While the problem of segmenting lumbar vertebrae has been addressed repeatedly in literature, the associated problem of computing relevant biomarkers on the basis of the segmentation has not been addressed thoroughly. In this paper, we propose a Random-Forest based approach that learns to segment vertebral bodies in CT images followed by a biomarker evaluation framework that extracts vertebral heights and widths from the segmentations obtained. Our dataset consists of 15 CT sagittal scans obtained from General Electric Healthcare. Our main approach is divided into three parts: the first stage is image pre-processing which is used to correct for variations in illumination across all the images followed by preparing the foreground and background objects from images; the next stage is Machine Learning using Random-Forests, which distinguishes the interest-point vectors between foreground or background; and the last step is image post-processing, which is crucial to refine the results of classifier. The Dice coefficient was used as a statistical validation metric to evaluate the performance of our segmentations with an average value of 0.725 for our dataset.

Classification of melanoma lesions using sparse coded features and random forests

NASA Astrophysics Data System (ADS)

Rastgoo, Mojdeh; Lemaître, Guillaume; Morel, Olivier; Massich, Joan; Garcia, Rafael; Meriaudeau, Fabrice; Marzani, Franck; Sidibé, Désiré

2016-03-01

Malignant melanoma is the most dangerous type of skin cancer, yet it is the most treatable kind of cancer, conditioned by its early diagnosis which is a challenging task for clinicians and dermatologists. In this regard, CAD systems based on machine learning and image processing techniques are developed to differentiate melanoma lesions from benign and dysplastic nevi using dermoscopic images. Generally, these frameworks are composed of sequential processes: pre-processing, segmentation, and classification. This architecture faces mainly two challenges: (i) each process is complex with the need to tune a set of parameters, and is specific to a given dataset; (ii) the performance of each process depends on the previous one, and the errors are accumulated throughout the framework. In this paper, we propose a framework for melanoma classification based on sparse coding which does not rely on any pre-processing or lesion segmentation. Our framework uses Random Forests classifier and sparse representation of three features: SIFT, Hue and Opponent angle histograms, and RGB intensities. The experiments are carried out on the public PH2 dataset using a 10-fold cross-validation. The results show that SIFT sparse-coded feature achieves the highest performance with sensitivity and specificity of 100% and 90.3% respectively, with a dictionary size of 800 atoms and a sparsity level of 2. Furthermore, the descriptor based on RGB intensities achieves similar results with sensitivity and specificity of 100% and 71.3%, respectively for a smaller dictionary size of 100 atoms. In conclusion, dictionary learning techniques encode strong structures of dermoscopic images and provide discriminant descriptors.

The edge-preservation multi-classifier relearning framework for the classification of high-resolution remotely sensed imagery

NASA Astrophysics Data System (ADS)

Han, Xiaopeng; Huang, Xin; Li, Jiayi; Li, Yansheng; Yang, Michael Ying; Gong, Jianya

2018-04-01

In recent years, the availability of high-resolution imagery has enabled more detailed observation of the Earth. However, it is imperative to simultaneously achieve accurate interpretation and preserve the spatial details for the classification of such high-resolution data. To this aim, we propose the edge-preservation multi-classifier relearning framework (EMRF). This multi-classifier framework is made up of support vector machine (SVM), random forest (RF), and sparse multinomial logistic regression via variable splitting and augmented Lagrangian (LORSAL) classifiers, considering their complementary characteristics. To better characterize complex scenes of remote sensing images, relearning based on landscape metrics is proposed, which iteratively quantizes both the landscape composition and spatial configuration by the use of the initial classification results. In addition, a novel tri-training strategy is proposed to solve the over-smoothing effect of relearning by means of automatic selection of training samples with low classification certainties, which always distribute in or near the edge areas. Finally, EMRF flexibly combines the strengths of relearning and tri-training via the classification certainties calculated by the probabilistic output of the respective classifiers. It should be noted that, in order to achieve an unbiased evaluation, we assessed the classification accuracy of the proposed framework using both edge and non-edge test samples. The experimental results obtained with four multispectral high-resolution images confirm the efficacy of the proposed framework, in terms of both edge and non-edge accuracy.

Maximizing lipocalin prediction through balanced and diversified training set and decision fusion.

PubMed

Nath, Abhigyan; Subbiah, Karthikeyan

2015-12-01

Lipocalins are short in sequence length and perform several important biological functions. These proteins are having less than 20% sequence similarity among paralogs. Experimentally identifying them is an expensive and time consuming process. The computational methods based on the sequence similarity for allocating putative members to this family are also far elusive due to the low sequence similarity existing among the members of this family. Consequently, the machine learning methods become a viable alternative for their prediction by using the underlying sequence/structurally derived features as the input. Ideally, any machine learning based prediction method must be trained with all possible variations in the input feature vector (all the sub-class input patterns) to achieve perfect learning. A near perfect learning can be achieved by training the model with diverse types of input instances belonging to the different regions of the entire input space. Furthermore, the prediction performance can be improved through balancing the training set as the imbalanced data sets will tend to produce the prediction bias towards majority class and its sub-classes. This paper is aimed to achieve (i) the high generalization ability without any classification bias through the diversified and balanced training sets as well as (ii) enhanced the prediction accuracy by combining the results of individual classifiers with an appropriate fusion scheme. Instead of creating the training set randomly, we have first used the unsupervised Kmeans clustering algorithm to create diversified clusters of input patterns and created the diversified and balanced training set by selecting an equal number of patterns from each of these clusters. Finally, probability based classifier fusion scheme was applied on boosted random forest algorithm (which produced greater sensitivity) and K nearest neighbour algorithm (which produced greater specificity) to achieve the enhanced predictive performance than that of individual base classifiers. The performance of the learned models trained on Kmeans preprocessed training set is far better than the randomly generated training sets. The proposed method achieved a sensitivity of 90.6%, specificity of 91.4% and accuracy of 91.0% on the first test set and sensitivity of 92.9%, specificity of 96.2% and accuracy of 94.7% on the second blind test set. These results have established that diversifying training set improves the performance of predictive models through superior generalization ability and balancing the training set improves prediction accuracy. For smaller data sets, unsupervised Kmeans based sampling can be an effective technique to increase generalization than that of the usual random splitting method. Copyright © 2015 Elsevier Ltd. All rights reserved.

iDBPs: a web server for the identification of DNA binding proteins.

PubMed

Nimrod, Guy; Schushan, Maya; Szilágyi, András; Leslie, Christina; Ben-Tal, Nir

2010-03-01

The iDBPs server uses the three-dimensional (3D) structure of a query protein to predict whether it binds DNA. First, the algorithm predicts the functional region of the protein based on its evolutionary profile; the assumption is that large clusters of conserved residues are good markers of functional regions. Next, various characteristics of the predicted functional region as well as global features of the protein are calculated, such as the average surface electrostatic potential, the dipole moment and cluster-based amino acid conservation patterns. Finally, a random forests classifier is used to predict whether the query protein is likely to bind DNA and to estimate the prediction confidence. We have trained and tested the classifier on various datasets and shown that it outperformed related methods. On a dataset that reflects the fraction of DNA binding proteins (DBPs) in a proteome, the area under the ROC curve was 0.90. The application of the server to an updated version of the N-Func database, which contains proteins of unknown function with solved 3D-structure, suggested new putative DBPs for experimental studies. http://idbps.tau.ac.il/

Diagnosis of Tempromandibular Disorders Using Local Binary Patterns.

PubMed

Haghnegahdar, A A; Kolahi, S; Khojastepour, L; Tajeripour, F

2018-03-01

Temporomandibular joint disorder (TMD) might be manifested as structural changes in bone through modification, adaptation or direct destruction. We propose to use Local Binary Pattern (LBP) characteristics and histogram-oriented gradients on the recorded images as a diagnostic tool in TMD assessment. CBCT images of 66 patients (132 joints) with TMD and 66 normal cases (132 joints) were collected and 2 coronal cut prepared from each condyle, although images were limited to head of mandibular condyle. In order to extract features of images, first we use LBP and then histogram of oriented gradients. To reduce dimensionality, the linear algebra Singular Value Decomposition (SVD) is applied to the feature vectors matrix of all images. For evaluation, we used K nearest neighbor (K-NN), Support Vector Machine, Naïve Bayesian and Random Forest classifiers. We used Receiver Operating Characteristic (ROC) to evaluate the hypothesis. K nearest neighbor classifier achieves a very good accuracy (0.9242), moreover, it has desirable sensitivity (0.9470) and specificity (0.9015) results, when other classifiers have lower accuracy, sensitivity and specificity. We proposed a fully automatic approach to detect TMD using image processing techniques based on local binary patterns and feature extraction. K-NN has been the best classifier for our experiments in detecting patients from healthy individuals, by 92.42% accuracy, 94.70% sensitivity and 90.15% specificity. The proposed method can help automatically diagnose TMD at its initial stages.

RF-Phos: A Novel General Phosphorylation Site Prediction Tool Based on Random Forest.

PubMed

Ismail, Hamid D; Jones, Ahoi; Kim, Jung H; Newman, Robert H; Kc, Dukka B

2016-01-01

Protein phosphorylation is one of the most widespread regulatory mechanisms in eukaryotes. Over the past decade, phosphorylation site prediction has emerged as an important problem in the field of bioinformatics. Here, we report a new method, termed Random Forest-based Phosphosite predictor 2.0 (RF-Phos 2.0), to predict phosphorylation sites given only the primary amino acid sequence of a protein as input. RF-Phos 2.0, which uses random forest with sequence and structural features, is able to identify putative sites of phosphorylation across many protein families. In side-by-side comparisons based on 10-fold cross validation and an independent dataset, RF-Phos 2.0 compares favorably to other popular mammalian phosphosite prediction methods, such as PhosphoSVM, GPS2.1, and Musite.

The Zoning of Forest Fire Potential of Gulestan Province Forests Using Granular Computing and MODIS Images

NASA Astrophysics Data System (ADS)

Jalilzadeh Shadlouei, A.; Delavar, M. R.

2013-09-01

There are many vegetation in Iran. This is because of extent of Iran and its width. One of these vegetation is forest vegetation most prevalent in Northern provinces named Guilan, Mazandaran, Gulestan, Ardebil as well as East Azerbaijan. These forests are always threatened by natural forest fires so much so that there have been reports of tens of fires in recent years. Forest fires are one of the major environmental as well as economic, social and security concerns in the world causing much damages. According to climatology, forest fires are one of the important factors in the formation and dispersion of vegetation. Also, regarding the environment, forest fires cause the emission of considerable amounts of greenhouse gases, smoke and dust into the atmosphere which in turn causes the earth temperature to rise up and are unhealthy to humans, animals and vegetation. In agriculture droughts are the usual side effects of these fires. The causes of forest fires could be categorized as either Human or Natural Causes. Naturally, it is impossible to completely contain forest fires; however, areas with high potentials of fire could be designated and analysed to decrease the risk of fires. The zoning of forest fire potential is a multi-criteria problem always accompanied by inherent uncertainty like other multi-criteria problems. So far, various methods and algorithm for zoning hazardous areas via Remote Sensing (RS) and Geospatial Information System (GIS) have been offered. This paper aims at zoning forest fire potential of Gulestan Province of Iran forests utilizing Remote Sensing, Geospatial Information System, meteorological data, MODIS images and granular computing method. Granular computing is part of granular mathematical and one way of solving multi-criteria problems such forest fire potential zoning supervised by one expert or some experts , and it offers rules for classification with the least inconsistencies. On the basis of the experts' opinion, 6 determinative criterias contributing to forest fires have been designated as follows: vegetation (NDVI), slope, aspect, temperature, humidity and proximity to roadways. By applying these variables on several tentatively selected areas and formation information tables and producing granular decision tree and extraction of rules, the zoning rules (for the areas in question) were extracted. According to them the zoning of the entire area has been conducted. The zoned areas have been classified into 5 categories: high hazard, medium hazard (high), medium hazard (low), low hazard (high), low hazard (low). According to the map, the zoning of most of the areas fall into the low hazard (high) class while the least number of areas have been classified as low hazard (low). Comparing the forest fires in these regions in 2010 with the MODIS data base for forest fires, it is concluded that areas with high hazards of forest fire have been classified with a 64 percent precision. In other word 64 percent of pixels that are in high hazard classification are classified according to MODIS data base. Using this method we obtain a good range of Perception. Manager will reduce forest fire concern using precautionary proceeding on hazardous area.

Sustainable Forest Management Support Based on the Spatial Distribution of Fuels for Fire Management

Treesearch

José Germán Flores Garnica; Juan de Dios Benavides Solorio; David Arturo Moreno Gonzalez

2006-01-01

Fire behavior simulation is based mainly on the fuel model-concept. However, there are great difficulties to develop the corresponding maps, therefore it is suggested the generation of four fuel maps (1-hour, 10-hours, 100-hours and alive). These maps will allow a better definition of the spatial variation of forest fuels, even within a zone classified as a given fuel...

Estimating Unbiased Land Cover Change Areas In The Colombian Amazon Using Landsat Time Series And Statistical Inference Methods

NASA Astrophysics Data System (ADS)

Arevalo, P. A.; Olofsson, P.; Woodcock, C. E.

2017-12-01

Unbiased estimation of the areas of conversion between land categories ("activity data") and their uncertainty is crucial for providing more robust calculations of carbon emissions to the atmosphere, as well as their removals. This is particularly important for the REDD+ mechanism of UNFCCC where an economic compensation is tied to the magnitude and direction of such fluxes. Dense time series of Landsat data and statistical protocols are becoming an integral part of forest monitoring efforts, but there are relatively few studies in the tropics focused on using these methods to advance operational MRV systems (Monitoring, Reporting and Verification). We present the results of a prototype methodology for continuous monitoring and unbiased estimation of activity data that is compliant with the IPCC Approach 3 for representation of land. We used a break detection algorithm (Continuous Change Detection and Classification, CCDC) to fit pixel-level temporal segments to time series of Landsat data in the Colombian Amazon. The segments were classified using a Random Forest classifier to obtain annual maps of land categories between 2001 and 2016. Using these maps, a biannual stratified sampling approach was implemented and unbiased stratified estimators constructed to calculate area estimates with confidence intervals for each of the stable and change classes. Our results provide evidence of a decrease in primary forest as a result of conversion to pastures, as well as increase in secondary forest as pastures are abandoned and the forest allowed to regenerate. Estimating areas of other land transitions proved challenging because of their very small mapped areas compared to stable classes like forest, which corresponds to almost 90% of the study area. Implications on remote sensing data processing, sample allocation and uncertainty reduction are also discussed.

Fault Diagnosis Strategies for SOFC-Based Power Generation Plants

PubMed Central

Costamagna, Paola; De Giorgi, Andrea; Gotelli, Alberto; Magistri, Loredana; Moser, Gabriele; Sciaccaluga, Emanuele; Trucco, Andrea

2016-01-01

The success of distributed power generation by plants based on solid oxide fuel cells (SOFCs) is hindered by reliability problems that can be mitigated through an effective fault detection and isolation (FDI) system. However, the numerous operating conditions under which such plants can operate and the random size of the possible faults make identifying damaged plant components starting from the physical variables measured in the plant very difficult. In this context, we assess two classical FDI strategies (model-based with fault signature matrix and data-driven with statistical classification) and the combination of them. For this assessment, a quantitative model of the SOFC-based plant, which is able to simulate regular and faulty conditions, is used. Moreover, a hybrid approach based on the random forest (RF) classification method is introduced to address the discrimination of regular and faulty situations due to its practical advantages. Working with a common dataset, the FDI performances obtained using the aforementioned strategies, with different sets of monitored variables, are observed and compared. We conclude that the hybrid FDI strategy, realized by combining a model-based scheme with a statistical classifier, outperforms the other strategies. In addition, the inclusion of two physical variables that should be measured inside the SOFCs can significantly improve the FDI performance, despite the actual difficulty in performing such measurements. PMID:27556472

Mapping the montane cloud forest of Taiwan using 12 year MODIS-derived ground fog frequency data.

PubMed

Schulz, Hans Martin; Li, Ching-Feng; Thies, Boris; Chang, Shih-Chieh; Bendix, Jörg

2017-01-01

Up until now montane cloud forest (MCF) in Taiwan has only been mapped for selected areas of vegetation plots. This paper presents the first comprehensive map of MCF distribution for the entire island. For its creation, a Random Forest model was trained with vegetation plots from the National Vegetation Database of Taiwan that were classified as "MCF" or "non-MCF". This model predicted the distribution of MCF from a raster data set of parameters derived from a digital elevation model (DEM), Landsat channels and texture measures derived from them as well as ground fog frequency data derived from the Moderate Resolution Imaging Spectroradiometer. While the DEM parameters and Landsat data predicted much of the cloud forest's location, local deviations in the altitudinal distribution of MCF linked to the monsoonal influence as well as the Massenerhebung effect (causing MCF in atypically low altitudes) were only captured once fog frequency data was included. Therefore, our study suggests that ground fog data are most useful for accurately mapping MCF.

A serum protein-based algorithm for the detection of Alzheimer disease.

PubMed

O'Bryant, Sid E; Xiao, Guanghua; Barber, Robert; Reisch, Joan; Doody, Rachelle; Fairchild, Thomas; Adams, Perrie; Waring, Steven; Diaz-Arrastia, Ramon

2010-09-01

To develop an algorithm that separates patients with Alzheimer disease (AD) from controls. Longitudinal case-control study. The Texas Alzheimer's Research Consortium project. Patients  We analyzed serum protein-based multiplex biomarker data from 197 patients diagnosed with AD and 203 controls. Main Outcome Measure  The total sample was randomized equally into training and test sets and random forest methods were applied to the training set to create a biomarker risk score. The biomarker risk score had a sensitivity and specificity of 0.80 and 0.91, respectively, and an area under the curve of 0.91 in detecting AD. When age, sex, education, and APOE status were added to the algorithm, the sensitivity, specificity, and area under the curve were 0.94, 0.84, and 0.95, respectively. These initial data suggest that serum protein-based biomarkers can be combined with clinical information to accurately classify AD. A disproportionate number of inflammatory and vascular markers were weighted most heavily in the analyses. Additionally, these markers consistently distinguished cases from controls in significant analysis of microarray, logistic regression, and Wilcoxon analyses, suggesting the existence of an inflammatory-related endophenotype of AD that may provide targeted therapeutic opportunities for this subset of patients.

Semi-supervised manifold learning with affinity regularization for Alzheimer's disease identification using positron emission tomography imaging.

PubMed

Lu, Shen; Xia, Yong; Cai, Tom Weidong; Feng, David Dagan

2015-01-01

Dementia, Alzheimer's disease (AD) in particular is a global problem and big threat to the aging population. An image based computer-aided dementia diagnosis method is needed to providing doctors help during medical image examination. Many machine learning based dementia classification methods using medical imaging have been proposed and most of them achieve accurate results. However, most of these methods make use of supervised learning requiring fully labeled image dataset, which usually is not practical in real clinical environment. Using large amount of unlabeled images can improve the dementia classification performance. In this study we propose a new semi-supervised dementia classification method based on random manifold learning with affinity regularization. Three groups of spatial features are extracted from positron emission tomography (PET) images to construct an unsupervised random forest which is then used to regularize the manifold learning objective function. The proposed method, stat-of-the-art Laplacian support vector machine (LapSVM) and supervised SVM are applied to classify AD and normal controls (NC). The experiment results show that learning with unlabeled images indeed improves the classification performance. And our method outperforms LapSVM on the same dataset.

A learning scheme for reach to grasp movements: on EMG-based interfaces using task specific motion decoding models.

PubMed

Liarokapis, Minas V; Artemiadis, Panagiotis K; Kyriakopoulos, Kostas J; Manolakos, Elias S

2013-09-01

A learning scheme based on random forests is used to discriminate between different reach to grasp movements in 3-D space, based on the myoelectric activity of human muscles of the upper-arm and the forearm. Task specificity for motion decoding is introduced in two different levels: Subspace to move toward and object to be grasped. The discrimination between the different reach to grasp strategies is accomplished with machine learning techniques for classification. The classification decision is then used in order to trigger an EMG-based task-specific motion decoding model. Task specific models manage to outperform "general" models providing better estimation accuracy. Thus, the proposed scheme takes advantage of a framework incorporating both a classifier and a regressor that cooperate advantageously in order to split the task space. The proposed learning scheme can be easily used to a series of EMG-based interfaces that must operate in real time, providing data-driven capabilities for multiclass problems, that occur in everyday life complex environments.

Microscale photo interpretation of forest and nonforest land classes

NASA Technical Reports Server (NTRS)

Aldrich, R. C.; Greentree, W. J.

1972-01-01

Remote sensing of forest and nonforest land classes are discussed, using microscale photointerpretation. Results include: (1.) Microscale IR color photography can be interpreted within reasonable limits of error to estimate forest area. (2.) Forest interpretation is best on winter photography with 97 percent or better accuracy. (3.) Broad forest types can be classified on microscale photography. (4.) Active agricultural land is classified most accurately on early summer photography. (5.) Six percent of all nonforest observations were misclassified as forest.

Development of a five-year mortality model in systemic sclerosis patients by different analytical approaches.

PubMed

Beretta, Lorenzo; Santaniello, Alessandro; Cappiello, Francesca; Chawla, Nitesh V; Vonk, Madelon C; Carreira, Patricia E; Allanore, Yannick; Popa-Diaconu, D A; Cossu, Marta; Bertolotti, Francesca; Ferraccioli, Gianfranco; Mazzone, Antonino; Scorza, Raffaella

2010-01-01

Systemic sclerosis (SSc) is a multiorgan disease with high mortality rates. Several clinical features have been associated with poor survival in different populations of SSc patients, but no clear and reproducible prognostic model to assess individual survival prediction in scleroderma patients has ever been developed. We used Cox regression and three data mining-based classifiers (Naïve Bayes Classifier [NBC], Random Forests [RND-F] and logistic regression [Log-Reg]) to develop a robust and reproducible 5-year prognostic model. All the models were built and internally validated by means of 5-fold cross-validation on a population of 558 Italian SSc patients. Their predictive ability and capability of generalisation was then tested on an independent population of 356 patients recruited from 5 external centres and finally compared to the predictions made by two SSc domain experts on the same population. The NBC outperformed the Cox-based classifier and the other data mining algorithms after internal cross-validation (area under receiving operator characteristic curve, AUROC: NBC=0.759; RND-F=0.736; Log-Reg=0.754 and Cox= 0.724). The NBC had also a remarkable and better trade-off between sensitivity and specificity (e.g. Balanced accuracy, BA) than the Cox-based classifier, when tested on an independent population of SSc patients (BA: NBC=0.769, Cox=0.622). The NBC was also superior to domain experts in predicting 5-year survival in this population (AUROC=0.829 vs. AUROC=0.788 and BA=0.769 vs. BA=0.67). We provide a model to make consistent 5-year prognostic predictions in SSc patients. Its internal validity, as well as capability of generalisation and reduced uncertainty compared to human experts support its use at bedside. Available at: http://www.nd.edu/~nchawla/survival.xls.

LocFuse: human protein-protein interaction prediction via classifier fusion using protein localization information.

PubMed

Zahiri, Javad; Mohammad-Noori, Morteza; Ebrahimpour, Reza; Saadat, Samaneh; Bozorgmehr, Joseph H; Goldberg, Tatyana; Masoudi-Nejad, Ali

2014-12-01

Protein-protein interaction (PPI) detection is one of the central goals of functional genomics and systems biology. Knowledge about the nature of PPIs can help fill the widening gap between sequence information and functional annotations. Although experimental methods have produced valuable PPI data, they also suffer from significant limitations. Computational PPI prediction methods have attracted tremendous attentions. Despite considerable efforts, PPI prediction is still in its infancy in complex multicellular organisms such as humans. Here, we propose a novel ensemble learning method, LocFuse, which is useful in human PPI prediction. This method uses eight different genomic and proteomic features along with four types of different classifiers. The prediction performance of this classifier selection method was found to be considerably better than methods employed hitherto. This confirms the complex nature of the PPI prediction problem and also the necessity of using biological information for classifier fusion. The LocFuse is available at: http://lbb.ut.ac.ir/Download/LBBsoft/LocFuse. The results revealed that if we divide proteome space according to the cellular localization of proteins, then the utility of some classifiers in PPI prediction can be improved. Therefore, to predict the interaction for any given protein pair, we can select the most accurate classifier with regard to the cellular localization information. Based on the results, we can say that the importance of different features for PPI prediction varies between differently localized proteins; however in general, our novel features, which were extracted from position-specific scoring matrices (PSSMs), are the most important ones and the Random Forest (RF) classifier performs best in most cases. LocFuse was developed with a user-friendly graphic interface and it is freely available for Linux, Mac OSX and MS Windows operating systems. Copyright © 2014 Elsevier Inc. All rights reserved.

SVM feature selection based rotation forest ensemble classifiers to improve computer-aided diagnosis of Parkinson disease.

PubMed

Ozcift, Akin

2012-08-01

Parkinson disease (PD) is an age-related deterioration of certain nerve systems, which affects movement, balance, and muscle control of clients. PD is one of the common diseases which affect 1% of people older than 60 years. A new classification scheme based on support vector machine (SVM) selected features to train rotation forest (RF) ensemble classifiers is presented for improving diagnosis of PD. The dataset contains records of voice measurements from 31 people, 23 with PD and each record in the dataset is defined with 22 features. The diagnosis model first makes use of a linear SVM to select ten most relevant features from 22. As a second step of the classification model, six different classifiers are trained with the subset of features. Subsequently, at the third step, the accuracies of classifiers are improved by the utilization of RF ensemble classification strategy. The results of the experiments are evaluated using three metrics; classification accuracy (ACC), Kappa Error (KE) and Area under the Receiver Operating Characteristic (ROC) Curve (AUC). Performance measures of two base classifiers, i.e. KStar and IBk, demonstrated an apparent increase in PD diagnosis accuracy compared to similar studies in literature. After all, application of RF ensemble classification scheme improved PD diagnosis in 5 of 6 classifiers significantly. We, numerically, obtained about 97% accuracy in RF ensemble of IBk (a K-Nearest Neighbor variant) algorithm, which is a quite high performance for Parkinson disease diagnosis.

Comparison of Random Forest and Parametric Imputation Models for Imputing Missing Data Using MICE: A CALIBER Study

PubMed Central

Shah, Anoop D.; Bartlett, Jonathan W.; Carpenter, James; Nicholas, Owen; Hemingway, Harry

2014-01-01

Multivariate imputation by chained equations (MICE) is commonly used for imputing missing data in epidemiologic research. The “true” imputation model may contain nonlinearities which are not included in default imputation models. Random forest imputation is a machine learning technique which can accommodate nonlinearities and interactions and does not require a particular regression model to be specified. We compared parametric MICE with a random forest-based MICE algorithm in 2 simulation studies. The first study used 1,000 random samples of 2,000 persons drawn from the 10,128 stable angina patients in the CALIBER database (Cardiovascular Disease Research using Linked Bespoke Studies and Electronic Records; 2001–2010) with complete data on all covariates. Variables were artificially made “missing at random,” and the bias and efficiency of parameter estimates obtained using different imputation methods were compared. Both MICE methods produced unbiased estimates of (log) hazard ratios, but random forest was more efficient and produced narrower confidence intervals. The second study used simulated data in which the partially observed variable depended on the fully observed variables in a nonlinear way. Parameter estimates were less biased using random forest MICE, and confidence interval coverage was better. This suggests that random forest imputation may be useful for imputing complex epidemiologic data sets in which some patients have missing data. PMID:24589914

Comparison of random forest and parametric imputation models for imputing missing data using MICE: a CALIBER study.

PubMed

Shah, Anoop D; Bartlett, Jonathan W; Carpenter, James; Nicholas, Owen; Hemingway, Harry

2014-03-15

Multivariate imputation by chained equations (MICE) is commonly used for imputing missing data in epidemiologic research. The "true" imputation model may contain nonlinearities which are not included in default imputation models. Random forest imputation is a machine learning technique which can accommodate nonlinearities and interactions and does not require a particular regression model to be specified. We compared parametric MICE with a random forest-based MICE algorithm in 2 simulation studies. The first study used 1,000 random samples of 2,000 persons drawn from the 10,128 stable angina patients in the CALIBER database (Cardiovascular Disease Research using Linked Bespoke Studies and Electronic Records; 2001-2010) with complete data on all covariates. Variables were artificially made "missing at random," and the bias and efficiency of parameter estimates obtained using different imputation methods were compared. Both MICE methods produced unbiased estimates of (log) hazard ratios, but random forest was more efficient and produced narrower confidence intervals. The second study used simulated data in which the partially observed variable depended on the fully observed variables in a nonlinear way. Parameter estimates were less biased using random forest MICE, and confidence interval coverage was better. This suggests that random forest imputation may be useful for imputing complex epidemiologic data sets in which some patients have missing data.

Combining deep residual neural network features with supervised machine learning algorithms to classify diverse food image datasets.

PubMed

McAllister, Patrick; Zheng, Huiru; Bond, Raymond; Moorhead, Anne

2018-04-01

Obesity is increasing worldwide and can cause many chronic conditions such as type-2 diabetes, heart disease, sleep apnea, and some cancers. Monitoring dietary intake through food logging is a key method to maintain a healthy lifestyle to prevent and manage obesity. Computer vision methods have been applied to food logging to automate image classification for monitoring dietary intake. In this work we applied pretrained ResNet-152 and GoogleNet convolutional neural networks (CNNs), initially trained using ImageNet Large Scale Visual Recognition Challenge (ILSVRC) dataset with MatConvNet package, to extract features from food image datasets; Food 5K, Food-11, RawFooT-DB, and Food-101. Deep features were extracted from CNNs and used to train machine learning classifiers including artificial neural network (ANN), support vector machine (SVM), Random Forest, and Naive Bayes. Results show that using ResNet-152 deep features with SVM with RBF kernel can accurately detect food items with 99.4% accuracy using Food-5K validation food image dataset and 98.8% with Food-5K evaluation dataset using ANN, SVM-RBF, and Random Forest classifiers. Trained with ResNet-152 features, ANN can achieve 91.34%, 99.28% when applied to Food-11 and RawFooT-DB food image datasets respectively and SVM with RBF kernel can achieve 64.98% with Food-101 image dataset. From this research it is clear that using deep CNN features can be used efficiently for diverse food item image classification. The work presented in this research shows that pretrained ResNet-152 features provide sufficient generalisation power when applied to a range of food image classification tasks. Copyright © 2018 Elsevier Ltd. All rights reserved.

Machine learning methods for the classification of gliomas: Initial results using features extracted from MR spectroscopy.

PubMed

Ranjith, G; Parvathy, R; Vikas, V; Chandrasekharan, Kesavadas; Nair, Suresh

2015-04-01

With the advent of new imaging modalities, radiologists are faced with handling increasing volumes of data for diagnosis and treatment planning. The use of automated and intelligent systems is becoming essential in such a scenario. Machine learning, a branch of artificial intelligence, is increasingly being used in medical image analysis applications such as image segmentation, registration and computer-aided diagnosis and detection. Histopathological analysis is currently the gold standard for classification of brain tumors. The use of machine learning algorithms along with extraction of relevant features from magnetic resonance imaging (MRI) holds promise of replacing conventional invasive methods of tumor classification. The aim of the study is to classify gliomas into benign and malignant types using MRI data. Retrospective data from 28 patients who were diagnosed with glioma were used for the analysis. WHO Grade II (low-grade astrocytoma) was classified as benign while Grade III (anaplastic astrocytoma) and Grade IV (glioblastoma multiforme) were classified as malignant. Features were extracted from MR spectroscopy. The classification was done using four machine learning algorithms: multilayer perceptrons, support vector machine, random forest and locally weighted learning. Three of the four machine learning algorithms gave an area under ROC curve in excess of 0.80. Random forest gave the best performance in terms of AUC (0.911) while sensitivity was best for locally weighted learning (86.1%). The performance of different machine learning algorithms in the classification of gliomas is promising. An even better performance may be expected by integrating features extracted from other MR sequences. © The Author(s) 2015 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav.

Mapping Deforestation area in North Korea Using Phenology-based Multi-Index and Random Forest

NASA Astrophysics Data System (ADS)

Jin, Y.; Sung, S.; Lee, D. K.; Jeong, S.

2016-12-01

Forest ecosystem provides ecological benefits to both humans and wildlife. Growing global demand for food and fiber is accelerating the pressure on the forest ecosystem in whole world from agriculture and logging. In recently, North Korea lost almost 40 % of its forests to crop fields for food production and cut-down of forest for fuel woods between 1990 and 2015. It led to the increased damage caused by natural disasters and is known to be one of the most forest degraded areas in the world. The characteristic of forest landscape in North Korea is complex and heterogeneous, the major landscape types in the forest are hillside farm, unstocked forest, natural forest and plateau vegetation. Remote sensing can be used for the forest degradation mapping of a dynamic landscape at a broad scale of detail and spatial distribution. Confusion mostly occurred between hillside farmland and unstocked forest, but also between unstocked forest and forest. Most previous forest degradation that used focused on the classification of broad types such as deforests area and sand from the perspective of land cover classification. The objective of this study is using random forest for mapping degraded forest in North Korea by phenological based vegetation index derived from MODIS products, which has various environmental factors such as vegetation, soil and water at a regional scale for improving accuracy. The model created by random forest resulted in an overall accuracy was 91.44%. Class user's accuracy of hillside farmland and unstocked forest were 97.2% and 84%%, which indicate the degraded forest. Unstocked forest had relative low user accuracy due to misclassified hillside farmland and forest samples. Producer's accuracy of hillside farmland and unstocked forest were 85.2% and 93.3%, repectly. In this case hillside farmland had lower produce accuracy mainly due to confusion with field, unstocked forest and forest. Such a classification of degraded forest could supply essential information to decide the priority of forest management and restoration in degraded forest area.

A systematic comparison of different object-based classification techniques using high spatial resolution imagery in agricultural environments

NASA Astrophysics Data System (ADS)

Li, Manchun; Ma, Lei; Blaschke, Thomas; Cheng, Liang; Tiede, Dirk

2016-07-01

Geographic Object-Based Image Analysis (GEOBIA) is becoming more prevalent in remote sensing classification, especially for high-resolution imagery. Many supervised classification approaches are applied to objects rather than pixels, and several studies have been conducted to evaluate the performance of such supervised classification techniques in GEOBIA. However, these studies did not systematically investigate all relevant factors affecting the classification (segmentation scale, training set size, feature selection and mixed objects). In this study, statistical methods and visual inspection were used to compare these factors systematically in two agricultural case studies in China. The results indicate that Random Forest (RF) and Support Vector Machines (SVM) are highly suitable for GEOBIA classifications in agricultural areas and confirm the expected general tendency, namely that the overall accuracies decline with increasing segmentation scale. All other investigated methods except for RF and SVM are more prone to obtain a lower accuracy due to the broken objects at fine scales. In contrast to some previous studies, the RF classifiers yielded the best results and the k-nearest neighbor classifier were the worst results, in most cases. Likewise, the RF and Decision Tree classifiers are the most robust with or without feature selection. The results of training sample analyses indicated that the RF and adaboost. M1 possess a superior generalization capability, except when dealing with small training sample sizes. Furthermore, the classification accuracies were directly related to the homogeneity/heterogeneity of the segmented objects for all classifiers. Finally, it was suggested that RF should be considered in most cases for agricultural mapping.

ANALYSIS OF SAMPLING TECHNIQUES FOR IMBALANCED DATA: AN N=648 ADNI STUDY

PubMed Central

Dubey, Rashmi; Zhou, Jiayu; Wang, Yalin; Thompson, Paul M.; Ye, Jieping

2013-01-01

Many neuroimaging applications deal with imbalanced imaging data. For example, in Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset, the mild cognitive impairment (MCI) cases eligible for the study are nearly two times the Alzheimer’s disease (AD) patients for structural magnetic resonance imaging (MRI) modality and six times the control cases for proteomics modality. Constructing an accurate classifier from imbalanced data is a challenging task. Traditional classifiers that aim to maximize the overall prediction accuracy tend to classify all data into the majority class. In this paper, we study an ensemble system of feature selection and data sampling for the class imbalance problem. We systematically analyze various sampling techniques by examining the efficacy of different rates and types of undersampling, oversampling, and a combination of over and under sampling approaches. We thoroughly examine six widely used feature selection algorithms to identify significant biomarkers and thereby reduce the complexity of the data. The efficacy of the ensemble techniques is evaluated using two different classifiers including Random Forest and Support Vector Machines based on classification accuracy, area under the receiver operating characteristic curve (AUC), sensitivity, and specificity measures. Our extensive experimental results show that for various problem settings in ADNI, (1). a balanced training set obtained with K-Medoids technique based undersampling gives the best overall performance among different data sampling techniques and no sampling approach; and (2). sparse logistic regression with stability selection achieves competitive performance among various feature selection algorithms. Comprehensive experiments with various settings show that our proposed ensemble model of multiple undersampled datasets yields stable and promising results. PMID:24176869

Analysis of sampling techniques for imbalanced data: An n = 648 ADNI study.

PubMed

Dubey, Rashmi; Zhou, Jiayu; Wang, Yalin; Thompson, Paul M; Ye, Jieping

2014-02-15

Many neuroimaging applications deal with imbalanced imaging data. For example, in Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset, the mild cognitive impairment (MCI) cases eligible for the study are nearly two times the Alzheimer's disease (AD) patients for structural magnetic resonance imaging (MRI) modality and six times the control cases for proteomics modality. Constructing an accurate classifier from imbalanced data is a challenging task. Traditional classifiers that aim to maximize the overall prediction accuracy tend to classify all data into the majority class. In this paper, we study an ensemble system of feature selection and data sampling for the class imbalance problem. We systematically analyze various sampling techniques by examining the efficacy of different rates and types of undersampling, oversampling, and a combination of over and undersampling approaches. We thoroughly examine six widely used feature selection algorithms to identify significant biomarkers and thereby reduce the complexity of the data. The efficacy of the ensemble techniques is evaluated using two different classifiers including Random Forest and Support Vector Machines based on classification accuracy, area under the receiver operating characteristic curve (AUC), sensitivity, and specificity measures. Our extensive experimental results show that for various problem settings in ADNI, (1) a balanced training set obtained with K-Medoids technique based undersampling gives the best overall performance among different data sampling techniques and no sampling approach; and (2) sparse logistic regression with stability selection achieves competitive performance among various feature selection algorithms. Comprehensive experiments with various settings show that our proposed ensemble model of multiple undersampled datasets yields stable and promising results. © 2013 Elsevier Inc. All rights reserved.

Creation of a Digital Surface Model and Extraction of Coarse Woody Debris from Terrestrial Laser Scans in an Open Eucalypt Woodland

NASA Astrophysics Data System (ADS)

Muir, J.; Phinn, S. R.; Armston, J.; Scarth, P.; Eyre, T.

2014-12-01

Coarse woody debris (CWD) provides important habitat for many species and plays a vital role in nutrient cycling within an ecosystem. In addition, CWD makes an important contribution to forest biomass and fuel loads. Airborne or space based remote sensing instruments typically do not detect CWD beneath the forest canopy. Terrestrial laser scanning (TLS) provides a ground based method for three-dimensional (3-D) reconstruction of surface features and CWD. This research produced a 3-D reconstruction of the ground surface and automatically classified coarse woody debris from registered TLS scans. The outputs will be used to inform the development of a site-based index for the assessment of forest condition, and quantitative assessments of biomass and fuel loads. A survey grade terrestrial laser scanner (Riegl VZ400) was used to scan 13 positions, in an open eucalypt woodland site at Karawatha Forest Park, near Brisbane, Australia. Scans were registered, and a digital surface model (DSM) produced using an intensity threshold and an iterative morphological filter. The DSMs produced from single scans were compared to the registered multi-scan point cloud using standard error metrics including: Root Mean Squared Error (RMSE), Mean Squared Error (MSE), range, absolute error and signed error. In addition the DSM was compared to a Digital Elevation Model (DEM) produced from Airborne Laser Scanning (ALS). Coarse woody debris was subsequently classified from the DSM using laser pulse properties, including: width and amplitude, as well as point spatial relationships (e.g. nearest neighbour slope vectors). Validation of the coarse woody debris classification was completed using true-colour photographs co-registered to the TLS point cloud. The volume and length of the coarse woody debris was calculated from the classified point cloud. A representative network of TLS sites will allow for up-scaling to large area assessment using airborne or space based sensors to monitor forest condition, biomass and fuel loads.

Predictive value of initial FDG-PET features for treatment response and survival in esophageal cancer patients treated with chemo-radiation therapy using a random forest classifier.

PubMed

Desbordes, Paul; Ruan, Su; Modzelewski, Romain; Pineau, Pascal; Vauclin, Sébastien; Gouel, Pierrick; Michel, Pierre; Di Fiore, Frédéric; Vera, Pierre; Gardin, Isabelle

2017-01-01

In oncology, texture features extracted from positron emission tomography with 18-fluorodeoxyglucose images (FDG-PET) are of increasing interest for predictive and prognostic studies, leading to several tens of features per tumor. To select the best features, the use of a random forest (RF) classifier was investigated. Sixty-five patients with an esophageal cancer treated with a combined chemo-radiation therapy were retrospectively included. All patients underwent a pretreatment whole-body FDG-PET. The patients were followed for 3 years after the end of the treatment. The response assessment was performed 1 month after the end of the therapy. Patients were classified as complete responders and non-complete responders. Sixty-one features were extracted from medical records and PET images. First, Spearman's analysis was performed to eliminate correlated features. Then, the best predictive and prognostic subsets of features were selected using a RF algorithm. These results were compared to those obtained by a Mann-Whitney U test (predictive study) and a univariate Kaplan-Meier analysis (prognostic study). Among the 61 initial features, 28 were not correlated. From these 28 features, the best subset of complementary features found using the RF classifier to predict response was composed of 2 features: metabolic tumor volume (MTV) and homogeneity from the co-occurrence matrix. The corresponding predictive value (AUC = 0.836 ± 0.105, Se = 82 ± 9%, Sp = 91 ± 12%) was higher than the best predictive results found using the Mann-Whitney test: busyness from the gray level difference matrix (P < 0.0001, AUC = 0.810, Se = 66%, Sp = 88%). The best prognostic subset found using RF was composed of 3 features: MTV and 2 clinical features (WHO status and nutritional risk index) (AUC = 0.822 ± 0.059, Se = 79 ± 9%, Sp = 95 ± 6%), while no feature was significantly prognostic according to the Kaplan-Meier analysis. The RF classifier can improve predictive and prognostic values compared to the Mann-Whitney U test and the univariate Kaplan-Meier survival analysis when applied to several tens of features in a limited patient database.

Advancements in automated tissue segmentation pipeline for contrast-enhanced CT scans of adult and pediatric patients

NASA Astrophysics Data System (ADS)

Somasundaram, Elanchezhian; Kaufman, Robert; Brady, Samuel

2017-03-01

The development of a random forests machine learning technique is presented for fully-automated neck, chest, abdomen, and pelvis tissue segmentation of CT images using Trainable WEKA (Waikato Environment for Knowledge Analysis) Segmentation (TWS) plugin of FIJI (ImageJ, NIH). The use of a single classifier model to segment six tissue classes (lung, fat, muscle, solid organ, blood/contrast agent, bone) in the CT images is studied. An automated unbiased scheme to sample pixels from the training images and generate a balanced training dataset over the seven classes is also developed. Two independent training datasets are generated from a pool of 4 adult (>55 kg) and 3 pediatric patients (<=55 kg) with 7 manually contoured slices for each patient. Classifier training investigated 28 image filters comprising a total of 272 features. Highly correlated and insignificant features are eliminated using Correlated Feature Subset (CFS) selection with Best First Search (BFS) algorithms in WEKA. The 2 training models (from the 2 training datasets) had 74 and 71 input training features, respectively. The study also investigated the effect of varying the number of trees (25, 50, 100, and 200) in the random forest algorithm. The performance of the 2 classifier models are evaluated on inter-patient intra-slice, intrapatient inter-slice and inter-patient inter-slice test datasets. The Dice similarity coefficients (DSC) and confusion matrices are used to understand the performance of the classifiers across the tissue segments. The effect of number of features in the training input on the performance of the classifiers for tissue classes with less than optimal DSC values is also studied. The average DSC values for the two training models on the inter-patient intra-slice test data are: 0.98, 0.89, 0.87, 0.79, 0.68, and 0.84, for lung, fat, muscle, solid organ, blood/contrast agent, and bone, respectively. The study demonstrated that a robust segmentation accuracy for lung, muscle and fat tissue classes. For solid-organ, blood/contrast and bone, the performance of the segmentation pipeline improved significantly by using the advanced capabilities of WEKA. However, further improvements are needed to reduce the noise in the segmentation.

Computational Short-cutting the Big Data Classification Bottleneck: Using the MODIS Land Cover Product to Derive a Consistent 30 m Landsat Land Cover Product of the Conterminous United States

NASA Astrophysics Data System (ADS)

Zhang, H.; Roy, D. P.

2016-12-01

Classification is a fundamental process in remote sensing used to relate pixel values to land cover classes present on the surface. The state of the practice for large area land cover classification is to classify satellite time series metrics with a supervised (i.e., training data dependent) non-parametric classifier. Classification accuracy generally increases with training set size. However, training data collection is expensive and the optimal training distribution over large areas is unknown. The MODIS 500 m land cover product is available globally on an annual basis and so provides a potentially very large source of land cover training data. A novel methodology to classify large volume Landsat data using high quality training data derived automatically from the MODIS land cover product is demonstrated for all of the Conterminous United States (CONUS). The known misclassification accuracy of the MODIS land cover product and the scale difference between the 500 m MODIS and 30 m Landsat data are accommodated for by a novel MODIS product filtering, Landsat pixel selection, and iterative training approach to balance the proportion of local and CONUS training data used. Three years of global Web-enabled Landsat data (WELD) data for all of the CONUS are classified using a random forest classifier and the results assessed using random forest `out-of-bag' training samples. The global WELD data are corrected to surface nadir BRDF-Adjusted Reflectance and are defined in 158 × 158 km tiles in the same projection and nested to the MODIS land cover products. This reduces the need to pre-process the considerable Landsat data volume (more than 14,000 Landsat 5 and 7 scenes per year over the CONUS covering 11,000 million 30 m pixels). The methodology is implemented in a parallel manner on WELD tile by tile basis but provides a wall-to-wall seamless 30 m land cover product. Detailed tile and CONUS results are presented and the potential for global production using the recently available global WELD products are discussed.

SEGMA: An Automatic SEGMentation Approach for Human Brain MRI Using Sliding Window and Random Forests

PubMed Central

Serag, Ahmed; Wilkinson, Alastair G.; Telford, Emma J.; Pataky, Rozalia; Sparrow, Sarah A.; Anblagan, Devasuda; Macnaught, Gillian; Semple, Scott I.; Boardman, James P.

2017-01-01

Quantitative volumes from brain magnetic resonance imaging (MRI) acquired across the life course may be useful for investigating long term effects of risk and resilience factors for brain development and healthy aging, and for understanding early life determinants of adult brain structure. Therefore, there is an increasing need for automated segmentation tools that can be applied to images acquired at different life stages. We developed an automatic segmentation method for human brain MRI, where a sliding window approach and a multi-class random forest classifier were applied to high-dimensional feature vectors for accurate segmentation. The method performed well on brain MRI data acquired from 179 individuals, analyzed in three age groups: newborns (38–42 weeks gestational age), children and adolescents (4–17 years) and adults (35–71 years). As the method can learn from partially labeled datasets, it can be used to segment large-scale datasets efficiently. It could also be applied to different populations and imaging modalities across the life course. PMID:28163680

Classification of acoustic emission signals using wavelets and Random Forests : Application to localized corrosion

NASA Astrophysics Data System (ADS)

Morizet, N.; Godin, N.; Tang, J.; Maillet, E.; Fregonese, M.; Normand, B.

2016-03-01

This paper aims to propose a novel approach to classify acoustic emission (AE) signals deriving from corrosion experiments, even if embedded into a noisy environment. To validate this new methodology, synthetic data are first used throughout an in-depth analysis, comparing Random Forests (RF) to the k-Nearest Neighbor (k-NN) algorithm. Moreover, a new evaluation tool called the alter-class matrix (ACM) is introduced to simulate different degrees of uncertainty on labeled data for supervised classification. Then, tests on real cases involving noise and crevice corrosion are conducted, by preprocessing the waveforms including wavelet denoising and extracting a rich set of features as input of the RF algorithm. To this end, a software called RF-CAM has been developed. Results show that this approach is very efficient on ground truth data and is also very promising on real data, especially for its reliability, performance and speed, which are serious criteria for the chemical industry.

Analysis of Radarsat-2 Full Polarimetric Data for Forest Mapping

NASA Astrophysics Data System (ADS)

Maghsoudi, Yasser

Forests are a major natural resource of the Earth and control a wide range of environmental processes. Forests comprise a major part of the planet's plant biodiversity and have an important role in the global hydrological and biochemical cycles. Among the numerous potential applications of remote sensing in forestry, forest mapping plays a vital role for characterization of the forest in terms of species. Particularly, in Canada where forests occupy 45% of the territory, representing more than 400 million hectares of the total Canadian continental area. In this thesis, the potential of polarimetric SAR (PolSAR) Radarsat-2 data for forest mapping is investigated. This thesis has two principle objectives. First is to propose algorithms for analyzing the PolSAR image data for forest mapping. There are a wide range of SAR parameters that can be derived from PolSAR data. In order to make full use of the discriminative power offered by all these parameters, two categories of methods are proposed. The methods are based on the concept of feature selection and classifier ensemble. First, a nonparametric definition of the evaluation function is proposed and hence the methods NFS and CBFS. Second, a fast wrapper algorithm is proposed for the evaluation function in feature selection and hence the methods FWFS and FWCBFS. Finally, to incorporate the neighboring pixels information in classification an extension of the FWCBFS method i.e. CCBFS is proposed. The second objective of this thesis is to provide a comparison between leaf-on (summer) and leaf-off (fall) season images for forest mapping. Two Radarsat-2 images acquired in fine quad-polarized mode were chosen for this study. The images were collected in leaf-on and leaf-off seasons. We also test the hypothesis whether combining the SAR parameters obtained from both images can provide better results than either individual datasets. The rationale for this combination is that every dataset has some parameters which may be useful for forest mapping. To assess the potential of the proposed methods their performance have been compared with each other and with the baseline classifiers. The baseline methods include the Wishart classifier, which is a commonly used classification method in PolSAR community, as well as an SVM classifier with the full set of parameters. Experimental results showed a better performance of the leaf-off image compared to that of leaf-on image for forest mapping. It is also shown that combining leaf-off parameters with leaf-on parameters can significantly improve the classification accuracy. Also, the classification results (in terms of the overall accuracy) compared to the baseline classifiers demonstrate the effectiveness of the proposed nonparametric scheme for forest mapping.

Cosmic string detection with tree-based machine learning

NASA Astrophysics Data System (ADS)

Vafaei Sadr, A.; Farhang, M.; Movahed, S. M. S.; Bassett, B.; Kunz, M.

2018-07-01

We explore the use of random forest and gradient boosting, two powerful tree-based machine learning algorithms, for the detection of cosmic strings in maps of the cosmic microwave background (CMB), through their unique Gott-Kaiser-Stebbins effect on the temperature anisotropies. The information in the maps is compressed into feature vectors before being passed to the learning units. The feature vectors contain various statistical measures of the processed CMB maps that boost cosmic string detectability. Our proposed classifiers, after training, give results similar to or better than claimed detectability levels from other methods for string tension, Gμ. They can make 3σ detection of strings with Gμ ≳ 2.1 × 10-10 for noise-free, 0.9'-resolution CMB observations. The minimum detectable tension increases to Gμ ≳ 3.0 × 10-8 for a more realistic, CMB S4-like (II) strategy, improving over previous results.

Cosmic String Detection with Tree-Based Machine Learning

NASA Astrophysics Data System (ADS)

Vafaei Sadr, A.; Farhang, M.; Movahed, S. M. S.; Bassett, B.; Kunz, M.

2018-05-01

We explore the use of random forest and gradient boosting, two powerful tree-based machine learning algorithms, for the detection of cosmic strings in maps of the cosmic microwave background (CMB), through their unique Gott-Kaiser-Stebbins effect on the temperature anisotropies. The information in the maps is compressed into feature vectors before being passed to the learning units. The feature vectors contain various statistical measures of the processed CMB maps that boost cosmic string detectability. Our proposed classifiers, after training, give results similar to or better than claimed detectability levels from other methods for string tension, Gμ. They can make 3σ detection of strings with Gμ ≳ 2.1 × 10-10 for noise-free, 0.9΄-resolution CMB observations. The minimum detectable tension increases to Gμ ≳ 3.0 × 10-8 for a more realistic, CMB S4-like (II) strategy, improving over previous results.

Machine learning methods in chemoinformatics

PubMed Central

Mitchell, John B O

2014-01-01

Machine learning algorithms are generally developed in computer science or adjacent disciplines and find their way into chemical modeling by a process of diffusion. Though particular machine learning methods are popular in chemoinformatics and quantitative structure–activity relationships (QSAR), many others exist in the technical literature. This discussion is methods-based and focused on some algorithms that chemoinformatics researchers frequently use. It makes no claim to be exhaustive. We concentrate on methods for supervised learning, predicting the unknown property values of a test set of instances, usually molecules, based on the known values for a training set. Particularly relevant approaches include Artificial Neural Networks, Random Forest, Support Vector Machine, k-Nearest Neighbors and naïve Bayes classifiers. WIREs Comput Mol Sci 2014, 4:468–481. How to cite this article: WIREs Comput Mol Sci 2014, 4:468–481. doi:10.1002/wcms.1183 PMID:25285160

See Change: Classifying single observation transients from HST using SNCosmo

NASA Astrophysics Data System (ADS)

Sofiatti Nunes, Caroline; Perlmutter, Saul; Nordin, Jakob; Rubin, David; Lidman, Chris; Deustua, Susana E.; Fruchter, Andrew S.; Aldering, Greg Scott; Brodwin, Mark; Cunha, Carlos E.; Eisenhardt, Peter R.; Gonzalez, Anthony H.; Jee, Myungkook J.; Hildebrandt, Hendrik; Hoekstra, Henk; Santos, Joana; Stanford, S. Adam; Stern, Dana R.; Fassbender, Rene; Richard, Johan; Rosati, Piero; Wechsler, Risa H.; Muzzin, Adam; Willis, Jon; Boehringer, Hans; Gladders, Michael; Goobar, Ariel; Amanullah, Rahman; Hook, Isobel; Huterer, Dragan; Huang, Jiasheng; Kim, Alex G.; Kowalski, Marek; Linder, Eric; Pain, Reynald; Saunders, Clare; Suzuki, Nao; Barbary, Kyle H.; Rykoff, Eli S.; Meyers, Joshua; Spadafora, Anthony L.; Hayden, Brian; Wilson, Gillian; Rozo, Eduardo; Hilton, Matt; Dixon, Samantha; Yen, Mike

2016-01-01

The Supernova Cosmology Project (SCP) is executing "See Change", a large HST program to look for possible variation in dark energy using supernovae at z>1. As part of the survey, we often must make time-critical follow-up decisions based on multicolor detection at a single epoch. We demonstrate the use of the SNCosmo software package to obtain simulated fluxes in the HST filters for type Ia and core-collapse supernovae at various redshifts. These simulations allow us to compare photometric data from HST with the distribution of the simulated SNe through methods such as Random Forest, a learning method for classification, and Gaussian Kernel Estimation. The results help us make informed decisions about triggered follow up using HST and ground based observatories to provide time-critical information needed about transients. Examples of this technique applied in the context of See Change are shown.

Object-Based Arctic Sea Ice Feature Extraction through High Spatial Resolution Aerial photos

NASA Astrophysics Data System (ADS)

Miao, X.; Xie, H.

2015-12-01

High resolution aerial photographs used to detect and classify sea ice features can provide accurate physical parameters to refine, validate, and improve climate models. However, manually delineating sea ice features, such as melt ponds, submerged ice, water, ice/snow, and pressure ridges, is time-consuming and labor-intensive. An object-based classification algorithm is developed to automatically extract sea ice features efficiently from aerial photographs taken during the Chinese National Arctic Research Expedition in summer 2010 (CHINARE 2010) in the MIZ near the Alaska coast. The algorithm includes four steps: (1) the image segmentation groups the neighboring pixels into objects based on the similarity of spectral and textural information; (2) the random forest classifier distinguishes four general classes: water, general submerged ice (GSI, including melt ponds and submerged ice), shadow, and ice/snow; (3) the polygon neighbor analysis separates melt ponds and submerged ice based on spatial relationship; and (4) pressure ridge features are extracted from shadow based on local illumination geometry. The producer's accuracy of 90.8% and user's accuracy of 91.8% are achieved for melt pond detection, and shadow shows a user's accuracy of 88.9% and producer's accuracies of 91.4%. Finally, pond density, pond fraction, ice floes, mean ice concentration, average ridge height, ridge profile, and ridge frequency are extracted from batch processing of aerial photos, and their uncertainties are estimated.

A multivariate study of mangrove morphology (Rhizophora mangle) using both above and below-water plant architecture

USGS Publications Warehouse

Brooks, R.A.; Bell, S.S.

2005-01-01

A descriptive study of the architecture of the red mangrove, Rhizophora mangle L., habitat of Tampa Bay, FL, was conducted to assess if plant architecture could be used to discriminate overwash from fringing forest type. Seven above-water (e.g., tree height, diameter at breast height, and leaf area) and 10 below-water (e.g., root density, root complexity, and maximum root order) architectural features were measured in eight mangrove stands. A multivariate technique (discriminant analysis) was used to test the ability of different models comprising above-water, below-water, or whole tree architecture to classify forest type. Root architectural features appear to be better than classical forestry measurements at discriminating between fringing and overwash forests but, regardless of the features loaded into the model, misclassification rates were high as forest type was only correctly classified in 66% of the cases. Based upon habitat architecture, the results of this study do not support a sharp distinction between overwash and fringing red mangrove forests in Tampa Bay but rather indicate that the two are architecturally undistinguishable. Therefore, within this northern portion of the geographic range of red mangroves, a more appropriate classification system based upon architecture may be one in which overwash and fringing forest types are combined into a single, "tide dominated" category. ?? 2005 Elsevier Ltd. All rights reserved.

Angular relational signature-based chest radiograph image view classification.

PubMed

Santosh, K C; Wendling, Laurent

2018-01-22

In a computer-aided diagnosis (CAD) system, especially for chest radiograph or chest X-ray (CXR) screening, CXR image view information is required. Automatically separating CXR image view, frontal and lateral can ease subsequent CXR screening process, since the techniques may not equally work for both views. We present a novel technique to classify frontal and lateral CXR images, where we introduce angular relational signature through force histogram to extract features and apply three different state-of-the-art classifiers: multi-layer perceptron, random forest, and support vector machine to make a decision. We validated our fully automatic technique on a set of 8100 images hosted by the U.S. National Library of Medicine (NLM), National Institutes of Health (NIH), and achieved an accuracy close to 100%. Our method outperforms the state-of-the-art methods in terms of processing time (less than or close to 2 s for the whole test data) while the accuracies can be compared, and therefore, it justifies its practicality. Graphical Abstract Interpreting chest X-ray (CXR) through the angular relational signature.

Understanding global climate change scenarios through bioclimate stratification

NASA Astrophysics Data System (ADS)

Soteriades, A. D.; Murray-Rust, D.; Trabucco, A.; Metzger, M. J.

2017-08-01

Despite progress in impact modelling, communicating and understanding the implications of climatic change projections is challenging due to inherent complexity and a cascade of uncertainty. In this letter, we present an alternative representation of global climate change projections based on shifts in 125 multivariate strata characterized by relatively homogeneous climate. These strata form climate analogues that help in the interpretation of climate change impacts. A Random Forests classifier was calculated and applied to 63 Coupled Model Intercomparison Project Phase 5 climate scenarios at 5 arcmin resolution. Results demonstrate how shifting bioclimate strata can summarize future environmental changes and form a middle ground, conveniently integrating current knowledge of climate change impact with the interpretation advantages of categorical data but with a level of detail that resembles a continuous surface at global and regional scales. Both the agreement in major change and differences between climate change projections are visually combined, facilitating the interpretation of complex uncertainty. By making the data and the classifier available we provide a climate service that helps facilitate communication and provide new insight into the consequences of climate change.

Prediction of Nucleotide Binding Peptides Using Star Graph Topological Indices.

PubMed

Liu, Yong; Munteanu, Cristian R; Fernández Blanco, Enrique; Tan, Zhiliang; Santos Del Riego, Antonino; Pazos, Alejandro

2015-11-01

The nucleotide binding proteins are involved in many important cellular processes, such as transmission of genetic information or energy transfer and storage. Therefore, the screening of new peptides for this biological function is an important research topic. The current study proposes a mixed methodology to obtain the first classification model that is able to predict new nucleotide binding peptides, using only the amino acid sequence. Thus, the methodology uses a Star graph molecular descriptor of the peptide sequences and the Machine Learning technique for the best classifier. The best model represents a Random Forest classifier based on two features of the embedded and non-embedded graphs. The performance of the model is excellent, considering similar models in the field, with an Area Under the Receiver Operating Characteristic Curve (AUROC) value of 0.938 and true positive rate (TPR) of 0.886 (test subset). The prediction of new nucleotide binding peptides with this model could be useful for drug target studies in drug development. © 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.

Automatic analysis of diabetic peripheral neuropathy using multi-scale quantitative morphology of nerve fibres in corneal confocal microscopy imaging.

PubMed

Dabbah, M A; Graham, J; Petropoulos, I N; Tavakoli, M; Malik, R A

2011-10-01

Diabetic peripheral neuropathy (DPN) is one of the most common long term complications of diabetes. Corneal confocal microscopy (CCM) image analysis is a novel non-invasive technique which quantifies corneal nerve fibre damage and enables diagnosis of DPN. This paper presents an automatic analysis and classification system for detecting nerve fibres in CCM images based on a multi-scale adaptive dual-model detection algorithm. The algorithm exploits the curvilinear structure of the nerve fibres and adapts itself to the local image information. Detected nerve fibres are then quantified and used as feature vectors for classification using random forest (RF) and neural networks (NNT) classifiers. We show, in a comparative study with other well known curvilinear detectors, that the best performance is achieved by the multi-scale dual model in conjunction with the NNT classifier. An evaluation of clinical effectiveness shows that the performance of the automated system matches that of ground-truth defined by expert manual annotation. Copyright © 2011 Elsevier B.V. All rights reserved.

Automatic Gleason grading of prostate cancer using SLIM and machine learning

NASA Astrophysics Data System (ADS)

Nguyen, Tan H.; Sridharan, Shamira; Marcias, Virgilia; Balla, Andre K.; Do, Minh N.; Popescu, Gabriel

2016-03-01

In this paper, we present an updated automatic diagnostic procedure for prostate cancer using quantitative phase imaging (QPI). In a recent report [1], we demonstrated the use of Random Forest for image segmentation on prostate cores imaged using QPI. Based on these label maps, we developed an algorithm to discriminate between regions with Gleason grade 3 and 4 prostate cancer in prostatectomy tissue. The Area-Under-Curve (AUC) of 0.79 for the Receiver Operating Curve (ROC) can be obtained for Gleason grade 4 detection in a binary classification between Grade 3 and Grade 4. Our dataset includes 280 benign cases and 141 malignant cases. We show that textural features in phase maps have strong diagnostic values since they can be used in combination with the label map to detect presence or absence of basal cells, which is a strong indicator for prostate carcinoma. A support vector machine (SVM) classifier trained on this new feature vector can classify cancer/non-cancer with an error rate of 0.23 and an AUC value of 0.83.

Automatic Screening and Grading of Age-Related Macular Degeneration from Texture Analysis of Fundus Images

PubMed Central

Phan, Thanh Vân; Seoud, Lama; Chakor, Hadi; Cheriet, Farida

2016-01-01

Age-related macular degeneration (AMD) is a disease which causes visual deficiency and irreversible blindness to the elderly. In this paper, an automatic classification method for AMD is proposed to perform robust and reproducible assessments in a telemedicine context. First, a study was carried out to highlight the most relevant features for AMD characterization based on texture, color, and visual context in fundus images. A support vector machine and a random forest were used to classify images according to the different AMD stages following the AREDS protocol and to evaluate the features' relevance. Experiments were conducted on a database of 279 fundus images coming from a telemedicine platform. The results demonstrate that local binary patterns in multiresolution are the most relevant for AMD classification, regardless of the classifier used. Depending on the classification task, our method achieves promising performances with areas under the ROC curve between 0.739 and 0.874 for screening and between 0.469 and 0.685 for grading. Moreover, the proposed automatic AMD classification system is robust with respect to image quality. PMID:27190636

Heterogeneous data fusion and intelligent techniques embedded in a mobile application for real-time chronic disease management.

PubMed

Bellos, Christos; Papadopoulos, Athanassios; Rosso, Roberto; Fotiadis, Dimitrios I

2011-01-01

CHRONIOUS system is an integrated platform aiming at the management of chronic disease patients. One of the most important components of the system is a Decision Support System (DSS) that has been developed in a Smart Device (SD). This component decides on patient's current health status by combining several data, which are acquired either by wearable sensors or manually inputted by the patient or retrieved from the specific database. In case no abnormal situation has been tracked, the DSS takes no action and remains deactivated until next abnormal situation pack of data are being acquired or next scheduled data being transmitted. The DSS that has been implemented is an integrated classification system with two parallel classifiers, combining an expert system (rule-based system) and a supervised classifier, such as Support Vector Machines (SVM), Random Forests, artificial Neural Networks (aNN like the Multi-Layer Perceptron), Decision Trees and Naïve Bayes. The above categorized system is useful for providing critical information about the health status of the patient.

Identification of COPD patients' health status using an intelligent system in the CHRONIOUS wearable platform.

PubMed

Bellos, Christos C; Papadopoulos, Athanasios; Rosso, Roberto; Fotiadis, Dimitrios I

2014-05-01

The CHRONIOUS system offers an integrated platform aiming at the effective management and real-time assessment of the health status of the patient suffering from chronic obstructive pulmonary disease (COPD). An intelligent system is developed for the analysis and the real-time evaluation of patient's condition. A hybrid classifier has been implemented on a personal digital assistant, combining a support vector machine, a random forest, and a rule-based system to provide a more advanced categorization scheme for the early and in real-time characterization of a COPD episode. This is followed by a severity estimation algorithm which classifies the identified pathological situation in different levels and triggers an alerting mechanism to provide an informative and instructive message/advice to the patient and the clinical supervisor. The system has been validated using data collected from 30 patients that have been annotated by experts indicating 1) the severity level of the current patient's health status, and 2) the COPD disease level of the recruited patients according to the GOLD guidelines. The achieved characterization accuracy has been found 94%.

Symbolic feature detection for image understanding

NASA Astrophysics Data System (ADS)

Aslan, Sinem; Akgül, Ceyhun Burak; Sankur, Bülent

2014-03-01

In this study we propose a model-driven codebook generation method used to assign probability scores to pixels in order to represent underlying local shapes they reside in. In the first version of the symbol library we limited ourselves to photometric and similarity transformations applied on eight prototypical shapes of flat plateau , ramp, valley, ridge, circular and elliptic respectively pit and hill and used randomized decision forest as the statistical classifier to compute shape class ambiguity of each pixel. We achieved90% accuracy in identification of known objects from alternate views, however, we could not outperform texture, global and local shape methods, but only color-based method in recognition of unknown objects. We present a progress plan to be accomplished as a future work to improve the proposed approach further.

Microscopic saw mark analysis: an empirical approach.

PubMed

Love, Jennifer C; Derrick, Sharon M; Wiersema, Jason M; Peters, Charles

2015-01-01

Microscopic saw mark analysis is a well published and generally accepted qualitative analytical method. However, little research has focused on identifying and mitigating potential sources of error associated with the method. The presented study proposes the use of classification trees and random forest classifiers as an optimal, statistically sound approach to mitigate the potential for error of variability and outcome error in microscopic saw mark analysis. The statistical model was applied to 58 experimental saw marks created with four types of saws. The saw marks were made in fresh human femurs obtained through anatomical gift and were analyzed using a Keyence digital microscope. The statistical approach weighed the variables based on discriminatory value and produced decision trees with an associated outcome error rate of 8.62-17.82%. © 2014 American Academy of Forensic Sciences.

Using deep learning for detecting gender in adult chest radiographs

NASA Astrophysics Data System (ADS)

Xue, Zhiyun; Antani, Sameer; Long, L. Rodney; Thoma, George R.

2018-03-01

In this paper, we present a method for automatically identifying the gender of an imaged person using their frontal chest x-ray images. Our work is motivated by the need to determine missing gender information in some datasets. The proposed method employs the technique of convolutional neural network (CNN) based deep learning and transfer learning to overcome the challenge of developing handcrafted features in limited data. Specifically, the method consists of four main steps: pre-processing, CNN feature extractor, feature selection, and classifier. The method is tested on a combined dataset obtained from several sources with varying acquisition quality resulting in different pre-processing steps that are applied for each. For feature extraction, we tested and compared four CNN architectures, viz., AlexNet, VggNet, GoogLeNet, and ResNet. We applied a feature selection technique, since the feature length is larger than the number of images. Two popular classifiers: SVM and Random Forest, are used and compared. We evaluated the classification performance by cross-validation and used seven performance measures. The best performer is the VggNet-16 feature extractor with the SVM classifier, with accuracy of 86.6% and ROC Area being 0.932 for 5-fold cross validation. We also discuss several misclassified cases and describe future work for performance improvement.

Change descriptors for determining nodule malignancy in national lung screening trial CT screening images

NASA Astrophysics Data System (ADS)

Geiger, Benjamin; Hawkins, Samuel; Hall, Lawrence O.; Goldgof, Dmitry B.; Balagurunathan, Yoganand; Gatenby, Robert A.; Gillies, Robert J.

2016-03-01

Pulmonary nodules are effectively diagnosed in CT scans, but determining their malignancy has been a challenge. The rate of change of the volume of a pulmonary nodule is known to be a prognostic factor for cancer development. In this study, we propose that other changes in imaging characteristics are similarly informative. We examined the combination of image features across multiple CT scans, taken from the National Lung Screening Trial, with individual scans of the same patient separated by approximately one year. By subtracting the values of existing features in multiple scans for the same patient, we were able to improve the ability of existing classification algorithms to determine whether a nodule will become malignant. We trained each classifier on 83 nodules determined to be malignant by biopsy and 172 nodules determined to be benign by their clinical stability through two years of no change; classifiers were tested on 77 malignant and 144 benign nodules, using a set of features that in a test-retest experiment were shown to be stable. An accuracy of 83.71% and AUC of 0.814 were achieved with the Random Forests classifier on a subset of features determined to be stable via test-retest reproducibility analysis, further reduced with the Correlation-based Feature Selection algorithm.

Comparative analysis of classification based algorithms for diabetes diagnosis using iris images.

PubMed

Samant, Piyush; Agarwal, Ravinder

2018-01-01

Photo-diagnosis is always an intriguing area for the researchers, with the advancement of image processing and computer machine vision techniques it have become more reliable and popular in recent years. The objective of this paper is to study the change in the features of iris, particularly irregularities in the pigmentation of certain areas of the iris with respect to diabetic health of an individual. Apart from the point that iris recognition concentrates on the overall structure of the iris, diagnostic techniques emphasises the local variations in the particular area of iris. Pre-image processing techniques have been applied to extract iris and thereafter, region of interest from the extracted iris have been cropped out. In order to observe the changes in the tissue pigmentation of region of interest, statistical, texture textural and wavelet features have been extracted. At the end, a comparison of accuracies of five different classifiers has been presented to classify two subject groups of diabetic and non-diabetic. Best classification accuracy has been calculated as 89.66% by the random forest classifier. Results have been shown the effectiveness and diagnostic significance of the proposed methodology. Presented piece of work offers a novel systemic perspective of non-invasive and automatic diabetic diagnosis.

Classification of Strawberry Fruit Shape by Machine Learning

NASA Astrophysics Data System (ADS)

Ishikawa, T.; Hayashi, A.; Nagamatsu, S.; Kyutoku, Y.; Dan, I.; Wada, T.; Oku, K.; Saeki, Y.; Uto, T.; Tanabata, T.; Isobe, S.; Kochi, N.

2018-05-01

Shape is one of the most important traits of agricultural products due to its relationships with the quality, quantity, and value of the products. For strawberries, the nine types of fruit shape were defined and classified by humans based on the sampler patterns of the nine types. In this study, we tested the classification of strawberry shapes by machine learning in order to increase the accuracy of the classification, and we introduce the concept of computerization into this field. Four types of descriptors were extracted from the digital images of strawberries: (1) the Measured Values (MVs) including the length of the contour line, the area, the fruit length and width, and the fruit width/length ratio; (2) the Ellipse Similarity Index (ESI); (3) Elliptic Fourier Descriptors (EFDs), and (4) Chain Code Subtraction (CCS). We used these descriptors for the classification test along with the random forest approach, and eight of the nine shape types were classified with combinations of MVs + CCS + EFDs. CCS is a descriptor that adds human knowledge to the chain codes, and it showed higher robustness in classification than the other descriptors. Our results suggest machine learning's high ability to classify fruit shapes accurately. We will attempt to increase the classification accuracy and apply the machine learning methods to other plant species.

Identification of DEP domain-containing proteins by a machine learning method and experimental analysis of their expression in human HCC tissues

NASA Astrophysics Data System (ADS)

Liao, Zhijun; Wang, Xinrui; Zeng, Yeting; Zou, Quan

2016-12-01

The Dishevelled/EGL-10/Pleckstrin (DEP) domain-containing (DEPDC) proteins have seven members. However, whether this superfamily can be distinguished from other proteins based only on the amino acid sequences, remains unknown. Here, we describe a computational method to segregate DEPDCs and non-DEPDCs. First, we examined the Pfam numbers of the known DEPDCs and used the longest sequences for each Pfam to construct a phylogenetic tree. Subsequently, we extracted 188-dimensional (188D) and 20D features of DEPDCs and non-DEPDCs and classified them with random forest classifier. We also mined the motifs of human DEPDCs to find the related domains. Finally, we designed experimental verification methods of human DEPDC expression at the mRNA level in hepatocellular carcinoma (HCC) and adjacent normal tissues. The phylogenetic analysis showed that the DEPDCs superfamily can be divided into three clusters. Moreover, the 188D and 20D features can both be used to effectively distinguish the two protein types. Motif analysis revealed that the DEP and RhoGAP domain was common in human DEPDCs, human HCC and the adjacent tissues that widely expressed DEPDCs. However, their regulation was not identical. In conclusion, we successfully constructed a binary classifier for DEPDCs and experimentally verified their expression in human HCC tissues.

Prediction of lysine ubiquitylation with ensemble classifier and feature selection.

PubMed

Zhao, Xiaowei; Li, Xiangtao; Ma, Zhiqiang; Yin, Minghao

2011-01-01

Ubiquitylation is an important process of post-translational modification. Correct identification of protein lysine ubiquitylation sites is of fundamental importance to understand the molecular mechanism of lysine ubiquitylation in biological systems. This paper develops a novel computational method to effectively identify the lysine ubiquitylation sites based on the ensemble approach. In the proposed method, 468 ubiquitylation sites from 323 proteins retrieved from the Swiss-Prot database were encoded into feature vectors by using four kinds of protein sequences information. An effective feature selection method was then applied to extract informative feature subsets. After different feature subsets were obtained by setting different starting points in the search procedure, they were used to train multiple random forests classifiers and then aggregated into a consensus classifier by majority voting. Evaluated by jackknife tests and independent tests respectively, the accuracy of the proposed predictor reached 76.82% for the training dataset and 79.16% for the test dataset, indicating that this predictor is a useful tool to predict lysine ubiquitylation sites. Furthermore, site-specific feature analysis was performed and it was shown that ubiquitylation is intimately correlated with the features of its surrounding sites in addition to features derived from the lysine site itself. The feature selection method is available upon request.

Early sinkhole detection using a drone-based thermal camera and image processing

NASA Astrophysics Data System (ADS)

Lee, Eun Ju; Shin, Sang Young; Ko, Byoung Chul; Chang, Chunho

2016-09-01

Accurate advance detection of the sinkholes that are occurring more frequently now is an important way of preventing human fatalities and property damage. Unlike naturally occurring sinkholes, human-induced ones in urban areas are typically due to groundwater disturbances and leaks of water and sewage caused by large-scale construction. Although many sinkhole detection methods have been developed, it is still difficult to predict sinkholes that occur in depth areas. In addition, conventional methods are inappropriate for scanning a large area because of their high cost. Therefore, this paper uses a drone combined with a thermal far-infrared (FIR) camera to detect potential sinkholes over a large area based on computer vision and pattern classification techniques. To make a standard dataset, we dug eight holes of depths 0.5-2 m in increments of 0.5 m and with a maximum width of 1 m. We filmed these using the drone-based FIR camera at a height of 50 m. We first detect candidate regions by analysing cold spots in the thermal images based on the fact that a sinkhole typically has a lower thermal energy than its background. Then, these regions are classified into sinkhole and non-sinkhole classes using a pattern classifier. In this study, we ensemble the classification results based on a light convolutional neural network (CNN) and those based on a Boosted Random Forest (BRF) with handcrafted features. We apply the proposed ensemble method successfully to sinkhole data for various sizes and depths in different environments, and prove that the CNN ensemble and the BRF one with handcrafted features are better at detecting sinkholes than other classifiers or standalone CNN.

A technique for extrapolating and validating forest cover across large regions. Calibrating AVHRR data with TM data

Treesearch

L.R. Iverson; E.A. Cook; R.L. Graham

1989-01-01

An approach to extending high-resolution forest cover information across large regions is presented and validated. Landsat Thematic Mapper (TM) data were classified into forest and nonforest for a portion of Jackson County, Illinois. The classified TM image was then used to determine the relationship between forest cover and the spectral signature of Advanced Very High...

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.

Predicting Health Care Utilization After Behavioral Health Referral Using Natural Language Processing and Machine Learning.

PubMed

Roysden, Nathaniel; Wright, Adam

2015-01-01

Mental health problems are an independent predictor of increased healthcare utilization. We created random forest classifiers for predicting two outcomes following a patient's first behavioral health encounter: decreased utilization by any amount (AUROC 0.74) and ultra-high absolute utilization (AUROC 0.88). These models may be used for clinical decision support by referring providers, to automatically detect patients who may benefit from referral, for cost management, or for risk/protection factor analysis.

Machine Learning Methods for Production Cases Analysis

NASA Astrophysics Data System (ADS)

Mokrova, Nataliya V.; Mokrov, Alexander M.; Safonova, Alexandra V.; Vishnyakov, Igor V.

2018-03-01

Approach to analysis of events occurring during the production process were proposed. Described machine learning system is able to solve classification tasks related to production control and hazard identification at an early stage. Descriptors of the internal production network data were used for training and testing of applied models. k-Nearest Neighbors and Random forest methods were used to illustrate and analyze proposed solution. The quality of the developed classifiers was estimated using standard statistical metrics, such as precision, recall and accuracy.

Learning in data-limited multimodal scenarios: Scandent decision forests and tree-based features.

PubMed

Hor, Soheil; Moradi, Mehdi

2016-12-01

Incomplete and inconsistent datasets often pose difficulties in multimodal studies. We introduce the concept of scandent decision trees to tackle these difficulties. Scandent trees are decision trees that optimally mimic the partitioning of the data determined by another decision tree, and crucially, use only a subset of the feature set. We show how scandent trees can be used to enhance the performance of decision forests trained on a small number of multimodal samples when we have access to larger datasets with vastly incomplete feature sets. Additionally, we introduce the concept of tree-based feature transforms in the decision forest paradigm. When combined with scandent trees, the tree-based feature transforms enable us to train a classifier on a rich multimodal dataset, and use it to classify samples with only a subset of features of the training data. Using this methodology, we build a model trained on MRI and PET images of the ADNI dataset, and then test it on cases with only MRI data. We show that this is significantly more effective in staging of cognitive impairments compared to a similar decision forest model trained and tested on MRI only, or one that uses other kinds of feature transform applied to the MRI data. Copyright © 2016. Published by Elsevier B.V.

An application of quantile random forests for predictive mapping of forest attributes

Treesearch

E.A. Freeman; G.G. Moisen

2015-01-01

Increasingly, random forest models are used in predictive mapping of forest attributes. Traditional random forests output the mean prediction from the random trees. Quantile regression forests (QRF) is an extension of random forests developed by Nicolai Meinshausen that provides non-parametric estimates of the median predicted value as well as prediction quantiles. It...

Spectroscopic diagnosis of laryngeal carcinoma using near-infrared Raman spectroscopy and random recursive partitioning ensemble techniques.

PubMed

Teh, Seng Khoon; Zheng, Wei; Lau, David P; Huang, Zhiwei

2009-06-01

In this work, we evaluated the diagnostic ability of near-infrared (NIR) Raman spectroscopy associated with the ensemble recursive partitioning algorithm based on random forests for identifying cancer from normal tissue in the larynx. A rapid-acquisition NIR Raman system was utilized for tissue Raman measurements at 785 nm excitation, and 50 human laryngeal tissue specimens (20 normal; 30 malignant tumors) were used for NIR Raman studies. The random forests method was introduced to develop effective diagnostic algorithms for classification of Raman spectra of different laryngeal tissues. High-quality Raman spectra in the range of 800-1800 cm(-1) can be acquired from laryngeal tissue within 5 seconds. Raman spectra differed significantly between normal and malignant laryngeal tissues. Classification results obtained from the random forests algorithm on tissue Raman spectra yielded a diagnostic sensitivity of 88.0% and specificity of 91.4% for laryngeal malignancy identification. The random forests technique also provided variables importance that facilitates correlation of significant Raman spectral features with cancer transformation. This study shows that NIR Raman spectroscopy in conjunction with random forests algorithm has a great potential for the rapid diagnosis and detection of malignant tumors in the larynx.

Landsat for practical forest type mapping - A test case

NASA Technical Reports Server (NTRS)

Bryant, E.; Dodge, A. G., Jr.; Warren, S. D.

1980-01-01

Computer classified Landsat maps are compared with a recent conventional inventory of forest lands in northern Maine. Over the 196,000 hectare area mapped, estimates of the areas of softwood, mixed wood and hardwood forest obtained by a supervised classification of the Landsat data and a standard inventory based on aerial photointerpretation, probability proportional to prediction, field sampling and a standard forest measurement program are found to agree to within 5%. The cost of the Landsat maps is estimated to be $0.065/hectare. It is concluded that satellite techniques are worth developing for forest inventories, although they are not yet refined enough to be incorporated into current practical inventories.

Prostate cancer prediction using the random forest algorithm that takes into account transrectal ultrasound findings, age, and serum levels of prostate-specific antigen.

PubMed

Xiao, Li-Hong; Chen, Pei-Ran; Gou, Zhong-Ping; Li, Yong-Zhong; Li, Mei; Xiang, Liang-Cheng; Feng, Ping

2017-01-01

The aim of this study is to evaluate the ability of the random forest algorithm that combines data on transrectal ultrasound findings, age, and serum levels of prostate-specific antigen to predict prostate carcinoma. Clinico-demographic data were analyzed for 941 patients with prostate diseases treated at our hospital, including age, serum prostate-specific antigen levels, transrectal ultrasound findings, and pathology diagnosis based on ultrasound-guided needle biopsy of the prostate. These data were compared between patients with and without prostate cancer using the Chi-square test, and then entered into the random forest model to predict diagnosis. Patients with and without prostate cancer differed significantly in age and serum prostate-specific antigen levels (P < 0.001), as well as in all transrectal ultrasound characteristics (P < 0.05) except uneven echo (P = 0.609). The random forest model based on age, prostate-specific antigen and ultrasound predicted prostate cancer with an accuracy of 83.10%, sensitivity of 65.64%, and specificity of 93.83%. Positive predictive value was 86.72%, and negative predictive value was 81.64%. By integrating age, prostate-specific antigen levels and transrectal ultrasound findings, the random forest algorithm shows better diagnostic performance for prostate cancer than either diagnostic indicator on its own. This algorithm may help improve diagnosis of the disease by identifying patients at high risk for biopsy.

An efficient ensemble learning method for gene microarray classification.

PubMed

Osareh, Alireza; Shadgar, Bita

2013-01-01

The gene microarray analysis and classification have demonstrated an effective way for the effective diagnosis of diseases and cancers. However, it has been also revealed that the basic classification techniques have intrinsic drawbacks in achieving accurate gene classification and cancer diagnosis. On the other hand, classifier ensembles have received increasing attention in various applications. Here, we address the gene classification issue using RotBoost ensemble methodology. This method is a combination of Rotation Forest and AdaBoost techniques which in turn preserve both desirable features of an ensemble architecture, that is, accuracy and diversity. To select a concise subset of informative genes, 5 different feature selection algorithms are considered. To assess the efficiency of the RotBoost, other nonensemble/ensemble techniques including Decision Trees, Support Vector Machines, Rotation Forest, AdaBoost, and Bagging are also deployed. Experimental results have revealed that the combination of the fast correlation-based feature selection method with ICA-based RotBoost ensemble is highly effective for gene classification. In fact, the proposed method can create ensemble classifiers which outperform not only the classifiers produced by the conventional machine learning but also the classifiers generated by two widely used conventional ensemble learning methods, that is, Bagging and AdaBoost.

Integrating in-situ, Landsat, and MODIS data for mapping in Southern African savannas: experiences of LCCS-based land-cover mapping in the Kalahari in Namibia.

PubMed

Hüttich, Christian; Herold, Martin; Strohbach, Ben J; Dech, Stefan

2011-05-01

Integrated ecosystem assessment initiatives are important steps towards a global biodiversity observing system. Reliable earth observation data are key information for tracking biodiversity change on various scales. Regarding the establishment of standardized environmental observation systems, a key question is: What can be observed on each scale and how can land cover information be transferred? In this study, a land cover map from a dry semi-arid savanna ecosystem in Namibia was obtained based on the UN LCCS, in-situ data, and MODIS and Landsat satellite imagery. In situ botanical relevé samples were used as baseline data for the definition of a standardized LCCS legend. A standard LCCS code for savanna vegetation types is introduced. An object-oriented segmentation of Landsat imagery was used as intermediate stage for downscaling in-situ training data on a coarse MODIS resolution. MODIS time series metrics of the growing season 2004/2005 were used to classify Kalahari vegetation types using a tree-based ensemble classifier (Random Forest). The prevailing Kalahari vegetation types based on LCCS was open broadleaved deciduous shrubland with an herbaceous layer which differs from the class assignments of the global and regional land-cover maps. The separability analysis based on Bhattacharya distance measurements applied on two LCCS levels indicated a relationship of spectral mapping dependencies of annual MODIS time series features due to the thematic detail of the classification scheme. The analysis of LCCS classifiers showed an increased significance of life-form composition and soil conditions to the mapping accuracy. An overall accuracy of 92.48% was achieved. Woody plant associations proved to be most stable due to small omission and commission errors. The case study comprised a first suitability assessment of the LCCS classifier approach for a southern African savanna ecosystem.

Differential privacy-based evaporative cooling feature selection and classification with relief-F and random forests.

PubMed

Le, Trang T; Simmons, W Kyle; Misaki, Masaya; Bodurka, Jerzy; White, Bill C; Savitz, Jonathan; McKinney, Brett A

2017-09-15

Classification of individuals into disease or clinical categories from high-dimensional biological data with low prediction error is an important challenge of statistical learning in bioinformatics. Feature selection can improve classification accuracy but must be incorporated carefully into cross-validation to avoid overfitting. Recently, feature selection methods based on differential privacy, such as differentially private random forests and reusable holdout sets, have been proposed. However, for domains such as bioinformatics, where the number of features is much larger than the number of observations p≫n , these differential privacy methods are susceptible to overfitting. We introduce private Evaporative Cooling, a stochastic privacy-preserving machine learning algorithm that uses Relief-F for feature selection and random forest for privacy preserving classification that also prevents overfitting. We relate the privacy-preserving threshold mechanism to a thermodynamic Maxwell-Boltzmann distribution, where the temperature represents the privacy threshold. We use the thermal statistical physics concept of Evaporative Cooling of atomic gases to perform backward stepwise privacy-preserving feature selection. On simulated data with main effects and statistical interactions, we compare accuracies on holdout and validation sets for three privacy-preserving methods: the reusable holdout, reusable holdout with random forest, and private Evaporative Cooling, which uses Relief-F feature selection and random forest classification. In simulations where interactions exist between attributes, private Evaporative Cooling provides higher classification accuracy without overfitting based on an independent validation set. In simulations without interactions, thresholdout with random forest and private Evaporative Cooling give comparable accuracies. We also apply these privacy methods to human brain resting-state fMRI data from a study of major depressive disorder. Code available at http://insilico.utulsa.edu/software/privateEC . brett-mckinney@utulsa.edu. Supplementary data are available at Bioinformatics online. © The Author (2017). Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com

Classification of the Gabon SAR Mosaic Using a Wavelet Based Rule Classifier

NASA Technical Reports Server (NTRS)

Simard, Marc; Saatchi, Sasan; DeGrandi, Gianfranco

2000-01-01

A method is developed for semi-automated classification of SAR images of the tropical forest. Information is extracted using the wavelet transform (WT). The transform allows for extraction of structural information in the image as a function of scale. In order to classify the SAR image, a Desicion Tree Classifier is used. The method of pruning is used to optimize classification rate versus tree size. The results give explicit insight on the type of information useful for a given class.

Prediction of recombinant protein overexpression in Escherichia coli using a machine learning based model (RPOLP).

PubMed

Habibi, Narjeskhatoon; Norouzi, Alireza; Mohd Hashim, Siti Z; Shamsir, Mohd Shahir; Samian, Razip

2015-11-01

Recombinant protein overexpression, an important biotechnological process, is ruled by complex biological rules which are mostly unknown, is in need of an intelligent algorithm so as to avoid resource-intensive lab-based trial and error experiments in order to determine the expression level of the recombinant protein. The purpose of this study is to propose a predictive model to estimate the level of recombinant protein overexpression for the first time in the literature using a machine learning approach based on the sequence, expression vector, and expression host. The expression host was confined to Escherichia coli which is the most popular bacterial host to overexpress recombinant proteins. To provide a handle to the problem, the overexpression level was categorized as low, medium and high. A set of features which were likely to affect the overexpression level was generated based on the known facts (e.g. gene length) and knowledge gathered from related literature. Then, a representative sub-set of features generated in the previous objective was determined using feature selection techniques. Finally a predictive model was developed using random forest classifier which was able to adequately classify the multi-class imbalanced small dataset constructed. The result showed that the predictive model provided a promising accuracy of 80% on average, in estimating the overexpression level of a recombinant protein. Copyright © 2015 Elsevier Ltd. All rights reserved.

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.

Multi-categorical deep learning neural network to classify retinal images: A pilot study employing small database.

PubMed

Choi, Joon Yul; Yoo, Tae Keun; Seo, Jeong Gi; Kwak, Jiyong; Um, Terry Taewoong; Rim, Tyler Hyungtaek

2017-01-01

Deep learning emerges as a powerful tool for analyzing medical images. Retinal disease detection by using computer-aided diagnosis from fundus image has emerged as a new method. We applied deep learning convolutional neural network by using MatConvNet for an automated detection of multiple retinal diseases with fundus photographs involved in STructured Analysis of the REtina (STARE) database. Dataset was built by expanding data on 10 categories, including normal retina and nine retinal diseases. The optimal outcomes were acquired by using a random forest transfer learning based on VGG-19 architecture. The classification results depended greatly on the number of categories. As the number of categories increased, the performance of deep learning models was diminished. When all 10 categories were included, we obtained results with an accuracy of 30.5%, relative classifier information (RCI) of 0.052, and Cohen's kappa of 0.224. Considering three integrated normal, background diabetic retinopathy, and dry age-related macular degeneration, the multi-categorical classifier showed accuracy of 72.8%, 0.283 RCI, and 0.577 kappa. In addition, several ensemble classifiers enhanced the multi-categorical classification performance. The transfer learning incorporated with ensemble classifier of clustering and voting approach presented the best performance with accuracy of 36.7%, 0.053 RCI, and 0.225 kappa in the 10 retinal diseases classification problem. First, due to the small size of datasets, the deep learning techniques in this study were ineffective to be applied in clinics where numerous patients suffering from various types of retinal disorders visit for diagnosis and treatment. Second, we found that the transfer learning incorporated with ensemble classifiers can improve the classification performance in order to detect multi-categorical retinal diseases. Further studies should confirm the effectiveness of algorithms with large datasets obtained from hospitals.

Exploring Google Earth Engine platform for big data processing: classification of multi-temporal satellite imagery for crop mapping

NASA Astrophysics Data System (ADS)

Shelestov, Andrii; Lavreniuk, Mykola; Kussul, Nataliia; Novikov, Alexei; Skakun, Sergii

2017-02-01

Many applied problems arising in agricultural monitoring and food security require reliable crop maps at national or global scale. Large scale crop mapping requires processing and management of large amount of heterogeneous satellite imagery acquired by various sensors that consequently leads to a “Big Data” problem. The main objective of this study is to explore efficiency of using the Google Earth Engine (GEE) platform when classifying multi-temporal satellite imagery with potential to apply the platform for a larger scale (e.g. country level) and multiple sensors (e.g. Landsat-8 and Sentinel-2). In particular, multiple state-of-the-art classifiers available in the GEE platform are compared to produce a high resolution (30 m) crop classification map for a large territory ( 28,100 km2 and 1.0 M ha of cropland). Though this study does not involve large volumes of data, it does address efficiency of the GEE platform to effectively execute complex workflows of satellite data processing required with large scale applications such as crop mapping. The study discusses strengths and weaknesses of classifiers, assesses accuracies that can be achieved with different classifiers for the Ukrainian landscape, and compares them to the benchmark classifier using a neural network approach that was developed in our previous studies. The study is carried out for the Joint Experiment of Crop Assessment and Monitoring (JECAM) test site in Ukraine covering the Kyiv region (North of Ukraine) in 2013. We found that Google Earth Engine (GEE) provides very good performance in terms of enabling access to the remote sensing products through the cloud platform and providing pre-processing; however, in terms of classification accuracy, the neural network based approach outperformed support vector machine (SVM), decision tree and random forest classifiers available in GEE.

Hip and Wrist Accelerometer Algorithms for Free-Living Behavior Classification.

PubMed

Ellis, Katherine; Kerr, Jacqueline; Godbole, Suneeta; Staudenmayer, John; Lanckriet, Gert

2016-05-01

Accelerometers are a valuable tool for objective measurement of physical activity (PA). Wrist-worn devices may improve compliance over standard hip placement, but more research is needed to evaluate their validity for measuring PA in free-living settings. Traditional cut-point methods for accelerometers can be inaccurate and need testing in free living with wrist-worn devices. In this study, we developed and tested the performance of machine learning (ML) algorithms for classifying PA types from both hip and wrist accelerometer data. Forty overweight or obese women (mean age = 55.2 ± 15.3 yr; BMI = 32.0 ± 3.7) wore two ActiGraph GT3X+ accelerometers (right hip, nondominant wrist; ActiGraph, Pensacola, FL) for seven free-living days. Wearable cameras captured ground truth activity labels. A classifier consisting of a random forest and hidden Markov model classified the accelerometer data into four activities (sitting, standing, walking/running, and riding in a vehicle). Free-living wrist and hip ML classifiers were compared with each other, with traditional accelerometer cut points, and with an algorithm developed in a laboratory setting. The ML classifier obtained average values of 89.4% and 84.6% balanced accuracy over the four activities using the hip and wrist accelerometer, respectively. In our data set with average values of 28.4 min of walking or running per day, the ML classifier predicted average values of 28.5 and 24.5 min of walking or running using the hip and wrist accelerometer, respectively. Intensity-based cut points and the laboratory algorithm significantly underestimated walking minutes. Our results demonstrate the superior performance of our PA-type classification algorithm, particularly in comparison with traditional cut points. Although the hip algorithm performed better, additional compliance achieved with wrist devices might justify using a slightly lower performing algorithm.

3D statistical shape models incorporating 3D random forest regression voting for robust CT liver segmentation

NASA Astrophysics Data System (ADS)

Norajitra, Tobias; Meinzer, Hans-Peter; Maier-Hein, Klaus H.

2015-03-01

During image segmentation, 3D Statistical Shape Models (SSM) usually conduct a limited search for target landmarks within one-dimensional search profiles perpendicular to the model surface. In addition, landmark appearance is modeled only locally based on linear profiles and weak learners, altogether leading to segmentation errors from landmark ambiguities and limited search coverage. We present a new method for 3D SSM segmentation based on 3D Random Forest Regression Voting. For each surface landmark, a Random Regression Forest is trained that learns a 3D spatial displacement function between the according reference landmark and a set of surrounding sample points, based on an infinite set of non-local randomized 3D Haar-like features. Landmark search is then conducted omni-directionally within 3D search spaces, where voxelwise forest predictions on landmark position contribute to a common voting map which reflects the overall position estimate. Segmentation experiments were conducted on a set of 45 CT volumes of the human liver, of which 40 images were randomly chosen for training and 5 for testing. Without parameter optimization, using a simple candidate selection and a single resolution approach, excellent results were achieved, while faster convergence and better concavity segmentation were observed, altogether underlining the potential of our approach in terms of increased robustness from distinct landmark detection and from better search coverage.

Thorough statistical comparison of machine learning regression models and their ensembles for sub-pixel imperviousness and imperviousness change mapping

NASA Astrophysics Data System (ADS)

Drzewiecki, Wojciech

2017-12-01

We evaluated the performance of nine machine learning regression algorithms and their ensembles for sub-pixel estimation of impervious areas coverages from Landsat imagery. The accuracy of imperviousness mapping in individual time points was assessed based on RMSE, MAE and R2. These measures were also used for the assessment of imperviousness change intensity estimations. The applicability for detection of relevant changes in impervious areas coverages at sub-pixel level was evaluated using overall accuracy, F-measure and ROC Area Under Curve. The results proved that Cubist algorithm may be advised for Landsat-based mapping of imperviousness for single dates. Stochastic gradient boosting of regression trees (GBM) may be also considered for this purpose. However, Random Forest algorithm is endorsed for both imperviousness change detection and mapping of its intensity. In all applications the heterogeneous model ensembles performed at least as well as the best individual models or better. They may be recommended for improving the quality of sub-pixel imperviousness and imperviousness change mapping. The study revealed also limitations of the investigated methodology for detection of subtle changes of imperviousness inside the pixel. None of the tested approaches was able to reliably classify changed and non-changed pixels if the relevant change threshold was set as one or three percent. Also for fi ve percent change threshold most of algorithms did not ensure that the accuracy of change map is higher than the accuracy of random classifi er. For the threshold of relevant change set as ten percent all approaches performed satisfactory.

Physical Human Activity Recognition Using Wearable Sensors.

PubMed

Attal, Ferhat; Mohammed, Samer; Dedabrishvili, Mariam; Chamroukhi, Faicel; Oukhellou, Latifa; Amirat, Yacine

2015-12-11

This paper presents a review of different classification techniques used to recognize human activities from wearable inertial sensor data. Three inertial sensor units were used in this study and were worn by healthy subjects at key points of upper/lower body limbs (chest, right thigh and left ankle). Three main steps describe the activity recognition process: sensors' placement, data pre-processing and data classification. Four supervised classification techniques namely, k-Nearest Neighbor (k-NN), Support Vector Machines (SVM), Gaussian Mixture Models (GMM), and Random Forest (RF) as well as three unsupervised classification techniques namely, k-Means, Gaussian mixture models (GMM) and Hidden Markov Model (HMM), are compared in terms of correct classification rate, F-measure, recall, precision, and specificity. Raw data and extracted features are used separately as inputs of each classifier. The feature selection is performed using a wrapper approach based on the RF algorithm. Based on our experiments, the results obtained show that the k-NN classifier provides the best performance compared to other supervised classification algorithms, whereas the HMM classifier is the one that gives the best results among unsupervised classification algorithms. This comparison highlights which approach gives better performance in both supervised and unsupervised contexts. It should be noted that the obtained results are limited to the context of this study, which concerns the classification of the main daily living human activities using three wearable accelerometers placed at the chest, right shank and left ankle of the subject.

Classification of optical coherence tomography images for diagnosing different ocular diseases

NASA Astrophysics Data System (ADS)

Gholami, Peyman; Sheikh Hassani, Mohsen; Kuppuswamy Parthasarathy, Mohana; Zelek, John S.; Lakshminarayanan, Vasudevan

2018-03-01

Optical Coherence tomography (OCT) images provide several indicators, e.g., the shape and the thickness of different retinal layers, which can be used for various clinical and non-clinical purposes. We propose an automated classification method to identify different ocular diseases, based on the local binary pattern features. The database consists of normal and diseased human eye SD-OCT images. We use a multiphase approach for building our classifier, including preprocessing, Meta learning, and active learning. Pre-processing is applied to the data to handle missing features from images and replace them with the mean or median of the corresponding feature. All the features are run through a Correlation-based Feature Subset Selection algorithm to detect the most informative features and omit the less informative ones. A Meta learning approach is applied to the data, in which a SVM and random forest are combined to obtain a more robust classifier. Active learning is also applied to strengthen our classifier around the decision boundary. The primary experimental results indicate that our method is able to differentiate between the normal and non-normal retina with an area under the ROC curve (AUC) of 98.6% and also to diagnose the three common retina-related diseases, i.e., Age-related Macular Degeneration, Diabetic Retinopathy, and Macular Hole, with an AUC of 100%, 95% and 83.8% respectively. These results indicate a better performance of the proposed method compared to most of the previous works in the literature.

Multi-factorial analysis of class prediction error: estimating optimal number of biomarkers for various classification rules.

PubMed

Khondoker, Mizanur R; Bachmann, Till T; Mewissen, Muriel; Dickinson, Paul; Dobrzelecki, Bartosz; Campbell, Colin J; Mount, Andrew R; Walton, Anthony J; Crain, Jason; Schulze, Holger; Giraud, Gerard; Ross, Alan J; Ciani, Ilenia; Ember, Stuart W J; Tlili, Chaker; Terry, Jonathan G; Grant, Eilidh; McDonnell, Nicola; Ghazal, Peter

2010-12-01

Machine learning and statistical model based classifiers have increasingly been used with more complex and high dimensional biological data obtained from high-throughput technologies. Understanding the impact of various factors associated with large and complex microarray datasets on the predictive performance of classifiers is computationally intensive, under investigated, yet vital in determining the optimal number of biomarkers for various classification purposes aimed towards improved detection, diagnosis, and therapeutic monitoring of diseases. We investigate the impact of microarray based data characteristics on the predictive performance for various classification rules using simulation studies. Our investigation using Random Forest, Support Vector Machines, Linear Discriminant Analysis and k-Nearest Neighbour shows that the predictive performance of classifiers is strongly influenced by training set size, biological and technical variability, replication, fold change and correlation between biomarkers. Optimal number of biomarkers for a classification problem should therefore be estimated taking account of the impact of all these factors. A database of average generalization errors is built for various combinations of these factors. The database of generalization errors can be used for estimating the optimal number of biomarkers for given levels of predictive accuracy as a function of these factors. Examples show that curves from actual biological data resemble that of simulated data with corresponding levels of data characteristics. An R package optBiomarker implementing the method is freely available for academic use from the Comprehensive R Archive Network (http://www.cran.r-project.org/web/packages/optBiomarker/).

Physical Human Activity Recognition Using Wearable Sensors

PubMed Central

Attal, Ferhat; Mohammed, Samer; Dedabrishvili, Mariam; Chamroukhi, Faicel; Oukhellou, Latifa; Amirat, Yacine

2015-01-01

This paper presents a review of different classification techniques used to recognize human activities from wearable inertial sensor data. Three inertial sensor units were used in this study and were worn by healthy subjects at key points of upper/lower body limbs (chest, right thigh and left ankle). Three main steps describe the activity recognition process: sensors’ placement, data pre-processing and data classification. Four supervised classification techniques namely, k-Nearest Neighbor (k-NN), Support Vector Machines (SVM), Gaussian Mixture Models (GMM), and Random Forest (RF) as well as three unsupervised classification techniques namely, k-Means, Gaussian mixture models (GMM) and Hidden Markov Model (HMM), are compared in terms of correct classification rate, F-measure, recall, precision, and specificity. Raw data and extracted features are used separately as inputs of each classifier. The feature selection is performed using a wrapper approach based on the RF algorithm. Based on our experiments, the results obtained show that the k-NN classifier provides the best performance compared to other supervised classification algorithms, whereas the HMM classifier is the one that gives the best results among unsupervised classification algorithms. This comparison highlights which approach gives better performance in both supervised and unsupervised contexts. It should be noted that the obtained results are limited to the context of this study, which concerns the classification of the main daily living human activities using three wearable accelerometers placed at the chest, right shank and left ankle of the subject. PMID:26690450

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.

Psoriasis skin biopsy image segmentation using Deep Convolutional Neural Network.

PubMed

Pal, Anabik; Garain, Utpal; Chandra, Aditi; Chatterjee, Raghunath; Senapati, Swapan

2018-06-01

Development of machine assisted tools for automatic analysis of psoriasis skin biopsy image plays an important role in clinical assistance. Development of automatic approach for accurate segmentation of psoriasis skin biopsy image is the initial prerequisite for developing such system. However, the complex cellular structure, presence of imaging artifacts, uneven staining variation make the task challenging. This paper presents a pioneering attempt for automatic segmentation of psoriasis skin biopsy images. Several deep neural architectures are tried for segmenting psoriasis skin biopsy images. Deep models are used for classifying the super-pixels generated by Simple Linear Iterative Clustering (SLIC) and the segmentation performance of these architectures is compared with the traditional hand-crafted feature based classifiers built on popularly used classifiers like K-Nearest Neighbor (KNN), Support Vector Machine (SVM) and Random Forest (RF). A U-shaped Fully Convolutional Neural Network (FCN) is also used in an end to end learning fashion where input is the original color image and the output is the segmentation class map for the skin layers. An annotated real psoriasis skin biopsy image data set of ninety (90) images is developed and used for this research. The segmentation performance is evaluated with two metrics namely, Jaccard's Coefficient (JC) and the Ratio of Correct Pixel Classification (RCPC) accuracy. The experimental results show that the CNN based approaches outperform the traditional hand-crafted feature based classification approaches. The present research shows that practical system can be developed for machine assisted analysis of psoriasis disease. Copyright © 2018 Elsevier B.V. All rights reserved.

Enhancing Critical Infrastructure and Key Resources (CIKR) Level-0 Physical Process Security Using Field Device Distinct Native Attribute Features

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

Lopez, Juan; Liefer, Nathan C.; Busho, Colin R.

Here, the need for improved Critical Infrastructure and Key Resource (CIKR) security is unquestioned and there has been minimal emphasis on Level-0 (PHY Process) improvements. Wired Signal Distinct Native Attribute (WS-DNA) Fingerprinting is investigated here as a non-intrusive PHY-based security augmentation to support an envisioned layered security strategy. Results are based on experimental response collections from Highway Addressable Remote Transducer (HART) Differential Pressure Transmitter (DPT) devices from three manufacturers (Yokogawa, Honeywell, Endress+Hauer) installed in an automated process control system. Device discrimination is assessed using Time Domain (TD) and Slope-Based FSK (SB-FSK) fingerprints input to Multiple Discriminant Analysis, Maximum Likelihood (MDA/ML)more » and Random Forest (RndF) classifiers. For 12 different classes (two devices per manufacturer at two distinct set points), both classifiers performed reliably and achieved an arbitrary performance benchmark of average cross-class percent correct of %C > 90%. The least challenging cross-manufacturer results included near-perfect %C ≈ 100%, while the more challenging like-model (serial number) discrimination results included 90%< %C < 100%, with TD Fingerprinting marginally outperforming SB-FSK Fingerprinting; SB-FSK benefits from having less stringent response alignment and registration requirements. The RndF classifier was most beneficial and enabled reliable selection of dimensionally reduced fingerprint subsets that minimize data storage and computational requirements. The RndF selected feature sets contained 15% of the full-dimensional feature sets and only suffered a worst case %CΔ = 3% to 4% performance degradation.« less

The Past, Present and Future of the Meteorological Phenomena Identification Near the Ground (mPING) Project

NASA Astrophysics Data System (ADS)

Elmore, K. L.

2016-12-01

The Metorological Phenomemna Identification NeartheGround (mPING) project is an example of a crowd-sourced, citizen science effort to gather data of sufficeint quality and quantity needed by new post processing methods that use machine learning. Transportation and infrastructure are particularly sensitive to precipitation type in winter weather. We extract attributes from operational numerical forecast models and use them in a random forest to generate forecast winter precipitation types. We find that random forests applied to forecast soundings are effective at generating skillful forecasts of surface ptype with consideralbly more skill than the current algorithms, especuially for ice pellets and freezing rain. We also find that three very different forecast models yuield similar overall results, showing that random forests are able to extract essentially equivalent information from different forecast models. We also show that the random forest for each model, and each profile type is unique to the particular forecast model and that the random forests developed using a particular model suffer significant degradation when given attributes derived from a different model. This implies that no single algorithm can perform well across all forecast models. Clearly, random forests extract information unavailable to "physically based" methods because the physical information in the models does not appear as we expect. One intersting result is that results from the classic "warm nose" sounding profile are, by far, the most sensitive to the particular forecast model, but this profile is also the one for which random forests are most skillful. Finally, a method for calibrarting probabilties for each different ptype using multinomial logistic regression is shown.

Comparison of Hyperspectral and Multispectral Satellites for Forest Alliance Classification in the San Francisco Bay Area

NASA Astrophysics Data System (ADS)

Clark, M. L.

2016-12-01

The goal of this study was to assess multi-temporal, Hyperspectral Infrared Imager (HyspIRI) satellite imagery for improved forest class mapping relative to multispectral satellites. The study area was the western San Francisco Bay Area, California and forest alliances (e.g., forest communities defined by dominant or co-dominant trees) were defined using the U.S. National Vegetation Classification System. Simulated 30-m HyspIRI, Landsat 8 and Sentinel-2 imagery were processed from image data acquired by NASA's AVIRIS airborne sensor in year 2015, with summer and multi-temporal (spring, summer, fall) data analyzed separately. HyspIRI reflectance was used to generate a suite of hyperspectral metrics that targeted key spectral features related to chemical and structural properties. The Random Forests classifier was applied to the simulated images and overall accuracies (OA) were compared to those from real Landsat 8 images. For each image group, broad land cover (e.g., Needle-leaf Trees, Broad-leaf Trees, Annual agriculture, Herbaceous, Built-up) was classified first, followed by a finer-detail forest alliance classification for pixels mapped as closed-canopy forest. There were 5 needle-leaf tree alliances and 16 broad-leaf tree alliances, including 7 Quercus (oak) alliance types. No forest alliance classification exceeded 50% OA, indicating that there was broad spectral similarity among alliances, most of which were not spectrally pure but rather a mix of tree species. In general, needle-leaf (Pine, Redwood, Douglas Fir) alliances had better class accuracies than broad-leaf alliances (Oaks, Madrone, Bay Laurel, Buckeye, etc). Multi-temporal data classifications all had 5-6% greater OA than with comparable summer data. For simulated data, HyspIRI metrics had 4-5% greater OA than Landsat 8 and Sentinel-2 multispectral imagery and 3-4% greater OA than HyspIRI reflectance. Finally, HyspIRI metrics had 8% greater OA than real Landsat 8 imagery. In conclusion, forest alliance classification was found to be a difficult remote sensing application with moderate resolution (30 m) satellite imagery; however, of the data tested, HyspIRI spectral metrics had the best performance relative to multispectral satellites.

Spectral Band Selection for Urban Material Classification Using Hyperspectral Libraries

NASA Astrophysics Data System (ADS)

Le Bris, A.; Chehata, N.; Briottet, X.; Paparoditis, N.

2016-06-01

In urban areas, information concerning very high resolution land cover and especially material maps are necessary for several city modelling or monitoring applications. That is to say, knowledge concerning the roofing materials or the different kinds of ground areas is required. Airborne remote sensing techniques appear to be convenient for providing such information at a large scale. However, results obtained using most traditional processing methods based on usual red-green-blue-near infrared multispectral images remain limited for such applications. A possible way to improve classification results is to enhance the imagery spectral resolution using superspectral or hyperspectral sensors. In this study, it is intended to design a superspectral sensor dedicated to urban materials classification and this work particularly focused on the selection of the optimal spectral band subsets for such sensor. First, reflectance spectral signatures of urban materials were collected from 7 spectral libraires. Then, spectral optimization was performed using this data set. The band selection workflow included two steps, optimising first the number of spectral bands using an incremental method and then examining several possible optimised band subsets using a stochastic algorithm. The same wrapper relevance criterion relying on a confidence measure of Random Forests classifier was used at both steps. To cope with the limited number of available spectra for several classes, additional synthetic spectra were generated from the collection of reference spectra: intra-class variability was simulated by multiplying reference spectra by a random coefficient. At the end, selected band subsets were evaluated considering the classification quality reached using a rbf svm classifier. It was confirmed that a limited band subset was sufficient to classify common urban materials. The important contribution of bands from the Short Wave Infra-Red (SWIR) spectral domain (1000-2400 nm) to material classification was also shown.

Performance evaluation of object based greenhouse detection from Sentinel-2 MSI and Landsat 8 OLI data: A case study from Almería (Spain)

NASA Astrophysics Data System (ADS)

Novelli, Antonio; Aguilar, Manuel A.; Nemmaoui, Abderrahim; Aguilar, Fernando J.; Tarantino, Eufemia

2016-10-01

This paper shows the first comparison between data from Sentinel-2 (S2) Multi Spectral Instrument (MSI) and Landsat 8 (L8) Operational Land Imager (OLI) headed up to greenhouse detection. Two closely related in time scenes, one for each sensor, were classified by using Object Based Image Analysis and Random Forest (RF). The RF input consisted of several object-based features computed from spectral bands and including mean values, spectral indices and textural features. S2 and L8 data comparisons were also extended using a common segmentation dataset extracted form VHR World-View 2 (WV2) imagery to test differences only due to their specific spectral contribution. The best band combinations to perform segmentation were found through a modified version of the Euclidian Distance 2 index. Four different RF classifications schemes were considered achieving 89.1%, 91.3%, 90.9% and 93.4% as the best overall accuracies respectively, evaluated over the whole study area.

Interpretation of fingerprint image quality features extracted by self-organizing maps

NASA Astrophysics Data System (ADS)

Danov, Ivan; Olsen, Martin A.; Busch, Christoph

2014-05-01

Accurate prediction of fingerprint quality is of significant importance to any fingerprint-based biometric system. Ensuring high quality samples for both probe and reference can substantially improve the system's performance by lowering false non-matches, thus allowing finer adjustment of the decision threshold of the biometric system. Furthermore, the increasing usage of biometrics in mobile contexts demands development of lightweight methods for operational environment. A novel two-tier computationally efficient approach was recently proposed based on modelling block-wise fingerprint image data using Self-Organizing Map (SOM) to extract specific ridge pattern features, which are then used as an input to a Random Forests (RF) classifier trained to predict the quality score of a propagated sample. This paper conducts an investigative comparative analysis on a publicly available dataset for the improvement of the two-tier approach by proposing additionally three feature interpretation methods, based respectively on SOM, Generative Topographic Mapping and RF. The analysis shows that two of the proposed methods produce promising results on the given dataset.

ReLiance: a machine learning and literature-based prioritization of receptor—ligand pairings

PubMed Central

Iacucci, Ernesto; Tranchevent, Léon-Charles; Popovic, Dusan; Pavlopoulos, Georgios A.; De Moor, Bart; Schneider, Reinhard; Moreau, Yves

2012-01-01

Motivation: The prediction of receptor—ligand pairings is an important area of research as intercellular communications are mediated by the successful interaction of these key proteins. As the exhaustive assaying of receptor—ligand pairs is impractical, a computational approach to predict pairings is necessary. We propose a workflow to carry out this interaction prediction task, using a text mining approach in conjunction with a state of the art prediction method, as well as a widely accessible and comprehensive dataset. Among several modern classifiers, random forests have been found to be the best at this prediction task. The training of this classifier was carried out using an experimentally validated dataset of Database of Ligand-Receptor Partners (DLRP) receptor—ligand pairs. New examples, co-cited with the training receptors and ligands, are then classified using the trained classifier. After applying our method, we find that we are able to successfully predict receptor—ligand pairs within the GPCR family with a balanced accuracy of 0.96. Upon further inspection, we find several supported interactions that were not present in the Database of Interacting Proteins (DIPdatabase). We have measured the balanced accuracy of our method resulting in high quality predictions stored in the available database ReLiance. Availability: http://homes.esat.kuleuven.be/~bioiuser/ReLianceDB/index.php Contact: yves.moreau@esat.kuleuven.be; ernesto.iacucci@gmail.com Supplementary information: Supplementary data are available at Bioinformatics online. PMID:22962483

Diagnosis of Tempromandibular Disorders Using Local Binary Patterns

PubMed Central

Haghnegahdar, A.A.; Kolahi, S.; Khojastepour, L.; Tajeripour, F.

2018-01-01

Background: Temporomandibular joint disorder (TMD) might be manifested as structural changes in bone through modification, adaptation or direct destruction. We propose to use Local Binary Pattern (LBP) characteristics and histogram-oriented gradients on the recorded images as a diagnostic tool in TMD assessment. Material and Methods: CBCT images of 66 patients (132 joints) with TMD and 66 normal cases (132 joints) were collected and 2 coronal cut prepared from each condyle, although images were limited to head of mandibular condyle. In order to extract features of images, first we use LBP and then histogram of oriented gradients. To reduce dimensionality, the linear algebra Singular Value Decomposition (SVD) is applied to the feature vectors matrix of all images. For evaluation, we used K nearest neighbor (K-NN), Support Vector Machine, Naïve Bayesian and Random Forest classifiers. We used Receiver Operating Characteristic (ROC) to evaluate the hypothesis. Results: K nearest neighbor classifier achieves a very good accuracy (0.9242), moreover, it has desirable sensitivity (0.9470) and specificity (0.9015) results, when other classifiers have lower accuracy, sensitivity and specificity. Conclusion: We proposed a fully automatic approach to detect TMD using image processing techniques based on local binary patterns and feature extraction. K-NN has been the best classifier for our experiments in detecting patients from healthy individuals, by 92.42% accuracy, 94.70% sensitivity and 90.15% specificity. The proposed method can help automatically diagnose TMD at its initial stages. PMID:29732343

Identification of DNA-Binding Proteins Using Structural, Electrostatic and Evolutionary Features

PubMed Central

Nimrod, Guy; Szilágyi, András; Leslie, Christina; Ben-Tal, Nir

2009-01-01

Summary DNA binding proteins (DBPs) often take part in various crucial processes of the cell's life cycle. Therefore, the identification and characterization of these proteins are of great importance. We present here a random forests classifier for identifying DBPs among proteins with known three-dimensional structures. First, clusters of evolutionarily conserved regions (patches) on the protein's surface are detected using the PatchFinder algorithm; previous studies showed that these regions are typically the proteins' functionally important regions. Next, we train a classifier using features like the electrostatic potential, cluster-based amino acid conservation patterns and the secondary structure content of the patches, as well as features of the whole protein including its dipole moment. Using 10-fold cross validation on a dataset of 138 DNA-binding proteins and 110 proteins which do not bind DNA, the classifier achieved a sensitivity and a specificity of 0.90, which is overall better than the performance of previously published methods. Furthermore, when we tested 5 different methods on 11 new DBPs which did not appear in the original dataset, only our method annotated all correctly. The resulting classifier was applied to a collection of 757 proteins of known structure and unknown function. Of these proteins, 218 were predicted to bind DNA, and we anticipate that some of them interact with DNA using new structural motifs. The use of complementary computational tools supports the notion that at least some of them do bind DNA. PMID:19233205

Identification of DNA-binding proteins using structural, electrostatic and evolutionary features.

PubMed

Nimrod, Guy; Szilágyi, András; Leslie, Christina; Ben-Tal, Nir

2009-04-10

DNA-binding proteins (DBPs) participate in various crucial processes in the life-cycle of the cells, and the identification and characterization of these proteins is of great importance. We present here a random forests classifier for identifying DBPs among proteins with known 3D structures. First, clusters of evolutionarily conserved regions (patches) on the surface of proteins were detected using the PatchFinder algorithm; earlier studies showed that these regions are typically the functionally important regions of proteins. Next, we trained a classifier using features like the electrostatic potential, cluster-based amino acid conservation patterns and the secondary structure content of the patches, as well as features of the whole protein, including its dipole moment. Using 10-fold cross-validation on a dataset of 138 DBPs and 110 proteins that do not bind DNA, the classifier achieved a sensitivity and a specificity of 0.90, which is overall better than the performance of published methods. Furthermore, when we tested five different methods on 11 new DBPs that did not appear in the original dataset, only our method annotated all correctly. The resulting classifier was applied to a collection of 757 proteins of known structure and unknown function. Of these proteins, 218 were predicted to bind DNA, and we anticipate that some of them interact with DNA using new structural motifs. The use of complementary computational tools supports the notion that at least some of them do bind DNA.

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.

Forest disturbance interactions and successional pathways in the Southern Rocky Mountains

USGS Publications Warehouse

Lu Liang,; Hawbaker, Todd J.; Zhu, Zhiliang; Xuecao Li,; Peng Gong,

2016-01-01

The pine forests in the southern portion of the Rocky Mountains are a heterogeneous mosaic of disturbance and recovery. The most extensive and intensive stress and mortality are received from human activity, fire, and mountain pine beetles (MPB;Dendroctonus ponderosae). Understanding disturbance interactions and disturbance-succession pathways are crucial for adapting management strategies to mitigate their impacts and anticipate future ecosystem change. Driven by this goal, we assessed the forest disturbance and recovery history in the Southern Rocky Mountains Ecoregion using a 13-year time series of Landsat image stacks. An automated classification workflow that integrates temporal segmentation techniques and a random forest classifier was used to examine disturbance patterns. To enhance efficiency in selecting representative samples at the ecoregion scale, a new sampling strategy that takes advantage of the scene-overlap among adjacent Landsat images was designed. The segment-based assessment revealed that the overall accuracy for all 14 scenes varied from 73.6% to 92.5%, with a mean of 83.1%. A design-based inference indicated the average producer’s and user’s accuracies for MPB mortality were 85.4% and 82.5% respectively. We found that burn severity was largely unrelated to the severity of pre-fire beetle outbreaks in this region, where the severity of post-fire beetle outbreaks generally decreased in relation to burn severity. Approximately half the clear-cut and burned areas were in various stages of recovery, but the regeneration rate was much slower for MPB-disturbed sites. Pre-fire beetle outbreaks and subsequent fire produced positive compound effects on seedling reestablishment in this ecoregion. Taken together, these results emphasize that although multiple disturbances do play a role in the resilience mechanism of the serotinous lodgepole pine, the overall recovery could be slow due to the vast area of beetle mortality.

Novel layered clustering-based approach for generating ensemble of classifiers.

PubMed

Rahman, Ashfaqur; Verma, Brijesh

2011-05-01

This paper introduces a novel concept for creating an ensemble of classifiers. The concept is based on generating an ensemble of classifiers through clustering of data at multiple layers. The ensemble classifier model generates a set of alternative clustering of a dataset at different layers by randomly initializing the clustering parameters and trains a set of base classifiers on the patterns at different clusters in different layers. A test pattern is classified by first finding the appropriate cluster at each layer and then using the corresponding base classifier. The decisions obtained at different layers are fused into a final verdict using majority voting. As the base classifiers are trained on overlapping patterns at different layers, the proposed approach achieves diversity among the individual classifiers. Identification of difficult-to-classify patterns through clustering as well as achievement of diversity through layering leads to better classification results as evidenced from the experimental results.

Forest management applications of Landsat data in a geographic information system

NASA Technical Reports Server (NTRS)

Maw, K. D.; Brass, J. A.

1982-01-01

The utility of land-cover data resulting from Landsat MSS classification can be greatly enhanced by use in combination with ancillary data. A demonstration forest management applications data base was constructed for Santa Cruz County, California, to demonstrate geographic information system applications of classified Landsat data. The data base contained detailed soils, digital terrain, land ownership, jurisdictional boundaries, fire events, and generalized land-use data, all registered to a UTM grid base. Applications models were developed from problems typical of fire management and reforestation planning.

Adjustments to forest inventory and analysis estimates of 2001 saw-log volumes for Kentucky

Treesearch

Stanley J. Zarnoch; Jeffery A. Turner

2005-01-01

The 2001 Kentucky Forest Inventory and Analysis survey overestimated hardwood saw-log volume in tree grade 1. This occurred because 2001 field crews classified too many trees as grade 1 trees. Data collected by quality assurance crews were used to generate two types of adjustments, one based on the proportion of trees misclassified and the other on the proportion of...

Effects of intermediate-scale wind disturbance on composition, structure, and succession in Quercus stands: Implications for natural disturbance-based silviculture

Treesearch

M.M. Cowden; J.L. Hart; C.J. Schweitzer; D.C. Dey

2014-01-01

Forest disturbances are discrete events in space and time that disrupt the biophysical environment and impart lasting legacies on forest composition and structure. Disturbances are often classified along a gradient of spatial extent and magnitude that ranges from catastrophic events where most of the overstory is removed to gap-scale events that modify local...

Critical Analysis of Forest Degradation in the Southern Eastern Ghats of India: Comparison of Satellite Imagery and Soil Quality Index

PubMed Central

Ramachandran, Andimuthu; Radhapriya, Parthasarathy; Jayakumar, Shanmuganathan; Dhanya, Praveen; Geetha, Rajadurai

2016-01-01

India has one of the largest assemblages of tropical biodiversity, with its unique floristic composition of endemic species. However, current forest cover assessment is performed via satellite-based forest surveys, which have many limitations. The present study, which was performed in the Eastern Ghats, analysed the satellite-based inventory provided by forest surveys and inferred from the results that this process no longer provides adequate information for quantifying forest degradation in an empirical manner. The study analysed 21 soil properties and generated a forest soil quality index of the Eastern Ghats, using principal component analysis. Using matrix modules and geospatial technology, we compared the forest degradation status calculated from satellite-based forest surveys with the degradation status calculated from the forest soil quality index. The Forest Survey of India classified about 1.8% of the Eastern Ghats’ total area as degraded forests and the remainder (98.2%) as open, dense, and very dense forests, whereas the soil quality index results found that about 42.4% of the total area is degraded, with the remainder (57.6%) being non-degraded. Our ground truth verification analyses indicate that the forest soil quality index along with the forest cover density data from the Forest Survey of India are ideal tools for evaluating forest degradation. PMID:26812397

An analysis of tree mortality using high resolution remotely-sensed data for mixed-conifer forests in San Diego county

NASA Astrophysics Data System (ADS)

Freeman, Mary Pyott

ABSTRACT An Analysis of Tree Mortality Using High Resolution Remotely-Sensed Data for Mixed-Conifer Forests in San Diego County by Mary Pyott Freeman The montane mixed-conifer forests of San Diego County are currently experiencing extensive tree mortality, which is defined as dieback where whole stands are affected. This mortality is likely the result of the complex interaction of many variables, such as altered fire regimes, climatic conditions such as drought, as well as forest pathogens and past management strategies. Conifer tree mortality and its spatial pattern and change over time were examined in three components. In component 1, two remote sensing approaches were compared for their effectiveness in delineating dead trees, a spatial contextual approach and an OBIA (object based image analysis) approach, utilizing various dates and spatial resolutions of airborne image data. For each approach transforms and masking techniques were explored, which were found to improve classifications, and an object-based assessment approach was tested. In component 2, dead tree maps produced by the most effective techniques derived from component 1 were utilized for point pattern and vector analyses to further understand spatio-temporal changes in tree mortality for the years 1997, 2000, 2002, and 2005 for three study areas: Palomar, Volcan and Laguna mountains. Plot-based fieldwork was conducted to further assess mortality patterns. Results indicate that conifer mortality was significantly clustered, increased substantially between 2002 and 2005, and was non-random with respect to tree species and diameter class sizes. In component 3, multiple environmental variables were used in Generalized Linear Model (GLM-logistic regression) and decision tree classifier model development, revealing the importance of climate and topographic factors such as precipitation and elevation, in being able to predict areas of high risk for tree mortality. The results from this study highlight the importance of multi-scale spatial as well as temporal analyses, in order to understand mixed-conifer forest structure, dynamics, and processes of decline, which can lead to more sustainable management of forests with continued natural and anthropogenic disturbance.

Relationship between structural features and water chemistry in boreal headwater streams--evaluation based on results from two water management survey tools suggested for Swedish forestry.

PubMed

Lestander, Ragna; Löfgren, Stefan; Henrikson, Lennart; Ågren, Anneli M

2015-04-01

Forestry may cause adverse impacts on water quality, and the forestry planning process is a key factor for the outcome of forest operation effects on stream water. To optimise environmental considerations and to identify actions needed to improve or maintain the stream biodiversity, two silvicultural water management tools, BIS+ (biodiversity, impact, sensitivity and added values) and Blue targeting, have been developed. In this study, we evaluate the links between survey variables, based on BIS+ and Blue targeting data, and water chemistry in 173 randomly selected headwater streams in the hemiboreal zone. While BIS+ and Blue targeting cannot replace more sophisticated monitoring methods necessary for classifying water quality in streams according to the EU Water Framework Directive (WFD, 2000/60/EC), our results lend support to the idea that the BIS+ protocol can be used to prioritise the protection of riparian forests. The relationship between BIS+ and water quality indicators (concentrations of nutrients and organic matter) together with data from fish studies suggests that this field protocol can be used to give reaches with higher biodiversity and conservation values a better protection. The tools indicate an ability to mitigate forestry impacts on water quality if the operations are adjusted to this knowledge in located areas.

Automation of motor dexterity assessment.

PubMed

Heyer, Patrick; Castrejon, Luis R; Orihuela-Espina, Felipe; Sucar, Luis Enrique

2017-07-01

Motor dexterity assessment is regularly performed in rehabilitation wards to establish patient status and automatization for such routinary task is sought. A system for automatizing the assessment of motor dexterity based on the Fugl-Meyer scale and with loose restrictions on sensing technologies is presented. The system consists of two main elements: 1) A data representation that abstracts the low level information obtained from a variety of sensors, into a highly separable low dimensionality encoding employing t-distributed Stochastic Neighbourhood Embedding, and, 2) central to this communication, a multi-label classifier that boosts classification rates by exploiting the fact that the classes corresponding to the individual exercises are naturally organized as a network. Depending on the targeted therapeutic movement class labels i.e. exercises scores, are highly correlated-patients who perform well in one, tends to perform well in related exercises-; and critically no node can be used as proxy of others - an exercise does not encode the information of other exercises. Over data from a cohort of 20 patients, the novel classifier outperforms classical Naive Bayes, random forest and variants of support vector machines (ANOVA: p < 0.001). The novel multi-label classification strategy fulfills an automatic system for motor dexterity assessment, with implications for lessening therapist's workloads, reducing healthcare costs and providing support for home-based virtual rehabilitation and telerehabilitation alternatives.

Temporally consistent probabilistic detection of new multiple sclerosis lesions in brain MRI.

PubMed

Elliott, Colm; Arnold, Douglas L; Collins, D Louis; Arbel, Tal

2013-08-01

Detection of new Multiple Sclerosis (MS) lesions on magnetic resonance imaging (MRI) is important as a marker of disease activity and as a potential surrogate for relapses. We propose an approach where sequential scans are jointly segmented, to provide a temporally consistent tissue segmentation while remaining sensitive to newly appearing lesions. The method uses a two-stage classification process: 1) a Bayesian classifier provides a probabilistic brain tissue classification at each voxel of reference and follow-up scans, and 2) a random-forest based lesion-level classification provides a final identification of new lesions. Generative models are learned based on 364 scans from 95 subjects from a multi-center clinical trial. The method is evaluated on sequential brain MRI of 160 subjects from a separate multi-center clinical trial, and is compared to 1) semi-automatically generated ground truth segmentations and 2) fully manual identification of new lesions generated independently by nine expert raters on a subset of 60 subjects. For new lesions greater than 0.15 cc in size, the classifier has near perfect performance (99% sensitivity, 2% false detection rate), as compared to ground truth. The proposed method was also shown to exceed the performance of any one of the nine expert manual identifications.

iDBPs: a web server for the identification of DNA binding proteins

PubMed Central

Nimrod, Guy; Schushan, Maya; Szilágyi, András; Leslie, Christina; Ben-Tal, Nir

2010-01-01

Summary: The iDBPs server uses the three-dimensional (3D) structure of a query protein to predict whether it binds DNA. First, the algorithm predicts the functional region of the protein based on its evolutionary profile; the assumption is that large clusters of conserved residues are good markers of functional regions. Next, various characteristics of the predicted functional region as well as global features of the protein are calculated, such as the average surface electrostatic potential, the dipole moment and cluster-based amino acid conservation patterns. Finally, a random forests classifier is used to predict whether the query protein is likely to bind DNA and to estimate the prediction confidence. We have trained and tested the classifier on various datasets and shown that it outperformed related methods. On a dataset that reflects the fraction of DNA binding proteins (DBPs) in a proteome, the area under the ROC curve was 0.90. The application of the server to an updated version of the N-Func database, which contains proteins of unknown function with solved 3D-structure, suggested new putative DBPs for experimental studies. Availability: http://idbps.tau.ac.il/ Contact: NirB@tauex.tau.ac.il Supplementary information: Supplementary data are available at Bioinformatics online. PMID:20089514

Android Malware Classification Using K-Means Clustering Algorithm

NASA Astrophysics Data System (ADS)

Hamid, Isredza Rahmi A.; Syafiqah Khalid, Nur; Azma Abdullah, Nurul; Rahman, Nurul Hidayah Ab; Chai Wen, Chuah

2017-08-01

Malware was designed to gain access or damage a computer system without user notice. Besides, attacker exploits malware to commit crime or fraud. This paper proposed Android malware classification approach based on K-Means clustering algorithm. We evaluate the proposed model in terms of accuracy using machine learning algorithms. Two datasets were selected to demonstrate the practicing of K-Means clustering algorithms that are Virus Total and Malgenome dataset. We classify the Android malware into three clusters which are ransomware, scareware and goodware. Nine features were considered for each types of dataset such as Lock Detected, Text Detected, Text Score, Encryption Detected, Threat, Porn, Law, Copyright and Moneypak. We used IBM SPSS Statistic software for data classification and WEKA tools to evaluate the built cluster. The proposed K-Means clustering algorithm shows promising result with high accuracy when tested using Random Forest algorithm.

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.

Characterizing stand-level forest canopy cover and height using Landsat time series, samples of airborne LiDAR, and the Random Forest algorithm

NASA Astrophysics Data System (ADS)

Ahmed, Oumer S.; Franklin, Steven E.; Wulder, Michael A.; White, Joanne C.

2015-03-01

Many forest management activities, including the development of forest inventories, require spatially detailed forest canopy cover and height data. Among the various remote sensing technologies, LiDAR (Light Detection and Ranging) offers the most accurate and consistent means for obtaining reliable canopy structure measurements. A potential solution to reduce the cost of LiDAR data, is to integrate transects (samples) of LiDAR data with frequently acquired and spatially comprehensive optical remotely sensed data. Although multiple regression is commonly used for such modeling, often it does not fully capture the complex relationships between forest structure variables. This study investigates the potential of Random Forest (RF), a machine learning technique, to estimate LiDAR measured canopy structure using a time series of Landsat imagery. The study is implemented over a 2600 ha area of industrially managed coastal temperate forests on Vancouver Island, British Columbia, Canada. We implemented a trajectory-based approach to time series analysis that generates time since disturbance (TSD) and disturbance intensity information for each pixel and we used this information to stratify the forest land base into two strata: mature forests and young forests. Canopy cover and height for three forest classes (i.e. mature, young and mature and young (combined)) were modeled separately using multiple regression and Random Forest (RF) techniques. For all forest classes, the RF models provided improved estimates relative to the multiple regression models. The lowest validation error was obtained for the mature forest strata in a RF model (R2 = 0.88, RMSE = 2.39 m and bias = -0.16 for canopy height; R2 = 0.72, RMSE = 0.068% and bias = -0.0049 for canopy cover). This study demonstrates the value of using disturbance and successional history to inform estimates of canopy structure and obtain improved estimates of forest canopy cover and height using the RF algorithm.

A study on real-time low-quality content detection on Twitter from the users' perspective.

PubMed

Chen, Weiling; Yeo, Chai Kiat; Lau, Chiew Tong; Lee, Bu Sung

2017-01-01

Detection techniques of malicious content such as spam and phishing on Online Social Networks (OSN) are common with little attention paid to other types of low-quality content which actually impacts users' content browsing experience most. The aim of our work is to detect low-quality content from the users' perspective in real time. To define low-quality content comprehensibly, Expectation Maximization (EM) algorithm is first used to coarsely classify low-quality tweets into four categories. Based on this preliminary study, a survey is carefully designed to gather users' opinions on different categories of low-quality content. Both direct and indirect features including newly proposed features are identified to characterize all types of low-quality content. We then further combine word level analysis with the identified features and build a keyword blacklist dictionary to improve the detection performance. We manually label an extensive Twitter dataset of 100,000 tweets and perform low-quality content detection in real time based on the characterized significant features and word level analysis. The results of our research show that our method has a high accuracy of 0.9711 and a good F1 of 0.8379 based on a random forest classifier with real time performance in the detection of low-quality content in tweets. Our work therefore achieves a positive impact in improving user experience in browsing social media content.

A study on real-time low-quality content detection on Twitter from the users’ perspective

PubMed Central

Yeo, Chai Kiat; Lau, Chiew Tong; Lee, Bu Sung

2017-01-01

Detection techniques of malicious content such as spam and phishing on Online Social Networks (OSN) are common with little attention paid to other types of low-quality content which actually impacts users’ content browsing experience most. The aim of our work is to detect low-quality content from the users’ perspective in real time. To define low-quality content comprehensibly, Expectation Maximization (EM) algorithm is first used to coarsely classify low-quality tweets into four categories. Based on this preliminary study, a survey is carefully designed to gather users’ opinions on different categories of low-quality content. Both direct and indirect features including newly proposed features are identified to characterize all types of low-quality content. We then further combine word level analysis with the identified features and build a keyword blacklist dictionary to improve the detection performance. We manually label an extensive Twitter dataset of 100,000 tweets and perform low-quality content detection in real time based on the characterized significant features and word level analysis. The results of our research show that our method has a high accuracy of 0.9711 and a good F1 of 0.8379 based on a random forest classifier with real time performance in the detection of low-quality content in tweets. Our work therefore achieves a positive impact in improving user experience in browsing social media content. PMID:28793347

An unsupervised two-stage clustering approach for forest structure classification based on X-band InSAR data - A case study in complex temperate forest stands

NASA Astrophysics Data System (ADS)

Abdullahi, Sahra; Schardt, Mathias; Pretzsch, Hans

2017-05-01

Forest structure at stand level plays a key role for sustainable forest management, since the biodiversity, productivity, growth and stability of the forest can be positively influenced by managing its structural diversity. In contrast to field-based measurements, remote sensing techniques offer a cost-efficient opportunity to collect area-wide information about forest stand structure with high spatial and temporal resolution. Especially Interferometric Synthetic Aperture Radar (InSAR), which facilitates worldwide acquisition of 3d information independent from weather conditions and illumination, is convenient to capture forest stand structure. This study purposes an unsupervised two-stage clustering approach for forest structure classification based on height information derived from interferometric X-band SAR data which was performed in complex temperate forest stands of Traunstein forest (South Germany). In particular, a four dimensional input data set composed of first-order height statistics was non-linearly projected on a two-dimensional Self-Organizing Map, spatially ordered according to similarity (based on the Euclidean distance) in the first stage and classified using the k-means algorithm in the second stage. The study demonstrated that X-band InSAR data exhibits considerable capabilities for forest structure classification. Moreover, the unsupervised classification approach achieved meaningful and reasonable results by means of comparison to aerial imagery and LiDAR data.

Hyperspectral analysis of clay minerals

NASA Astrophysics Data System (ADS)

Janaki Rama Suresh, G.; Sreenivas, K.; Sivasamy, R.

2014-11-01

A study was carried out by collecting soil samples from parts of Gwalior and Shivpuri district, Madhya Pradesh in order to assess the dominant clay mineral of these soils using hyperspectral data, as 0.4 to 2.5 μm spectral range provides abundant and unique information about many important earth-surface minerals. Understanding the spectral response along with the soil chemical properties can provide important clues for retrieval of mineralogical soil properties. The soil samples were collected based on stratified random sampling approach and dominant clay minerals were identified through XRD analysis. The absorption feature parameters like depth, width, area and asymmetry of the absorption peaks were derived from spectral profile of soil samples through DISPEC tool. The derived absorption feature parameters were used as inputs for modelling the dominant soil clay mineral present in the unknown samples using Random forest approach which resulted in kappa accuracy of 0.795. Besides, an attempt was made to classify the Hyperion data using Spectral Angle Mapper (SAM) algorithm with an overall accuracy of 68.43 %. Results showed that kaolinite was the dominant mineral present in the soils followed by montmorillonite in the study area.

Layers: A molecular surface peeling algorithm and its applications to analyze protein structures

PubMed Central

Karampudi, Naga Bhushana Rao; Bahadur, Ranjit Prasad

2015-01-01

We present an algorithm ‘Layers’ to peel the atoms of proteins as layers. Using Layers we show an efficient way to transform protein structures into 2D pattern, named residue transition pattern (RTP), which is independent of molecular orientations. RTP explains the folding patterns of proteins and hence identification of similarity between proteins is simple and reliable using RTP than with the standard sequence or structure based methods. Moreover, Layers generates a fine-tunable coarse model for the molecular surface by using non-random sampling. The coarse model can be used for shape comparison, protein recognition and ligand design. Additionally, Layers can be used to develop biased initial configuration of molecules for protein folding simulations. We have developed a random forest classifier to predict the RTP of a given polypeptide sequence. Layers is a standalone application; however, it can be merged with other applications to reduce the computational load when working with large datasets of protein structures. Layers is available freely at http://www.csb.iitkgp.ernet.in/applications/mol_layers/main. PMID:26553411

Vegetation survey in Amazonia using LANDSAT data. [Brazil

NASA Technical Reports Server (NTRS)

Parada, N. D. J. (Principal Investigator); Shimabukuro, Y. E.; Dossantos, J. R.; Deaquino, L. C. S.

1982-01-01

Automatic Image-100 analysis of LANDSAT data was performed using the MAXVER classification algorithm. In the pilot area, four vegetation units were mapped automatically in addition to the areas occupied for agricultural activities. The Image-100 classified results together with a soil map and information from RADAR images, permitted the establishment of the final legend with six classes: semi-deciduous tropical forest; low land evergreen tropical forest; secondary vegetation; tropical forest of humid areas, predominant pastureland and flood plains. Two water types were identified based on their sediments indicating different geological and geomorphological aspects.

Optimal Symmetric Multimodal Templates and Concatenated Random Forests for Supervised Brain Tumor Segmentation (Simplified) with ANTsR.

PubMed

Tustison, Nicholas J; Shrinidhi, K L; Wintermark, Max; Durst, Christopher R; Kandel, Benjamin M; Gee, James C; Grossman, Murray C; Avants, Brian B

2015-04-01

Segmenting and quantifying gliomas from MRI is an important task for diagnosis, planning intervention, and for tracking tumor changes over time. However, this task is complicated by the lack of prior knowledge concerning tumor location, spatial extent, shape, possible displacement of normal tissue, and intensity signature. To accommodate such complications, we introduce a framework for supervised segmentation based on multiple modality intensity, geometry, and asymmetry feature sets. These features drive a supervised whole-brain and tumor segmentation approach based on random forest-derived probabilities. The asymmetry-related features (based on optimal symmetric multimodal templates) demonstrate excellent discriminative properties within this framework. We also gain performance by generating probability maps from random forest models and using these maps for a refining Markov random field regularized probabilistic segmentation. This strategy allows us to interface the supervised learning capabilities of the random forest model with regularized probabilistic segmentation using the recently developed ANTsR package--a comprehensive statistical and visualization interface between the popular Advanced Normalization Tools (ANTs) and the R statistical project. The reported algorithmic framework was the top-performing entry in the MICCAI 2013 Multimodal Brain Tumor Segmentation challenge. The challenge data were widely varying consisting of both high-grade and low-grade glioma tumor four-modality MRI from five different institutions. Average Dice overlap measures for the final algorithmic assessment were 0.87, 0.78, and 0.74 for "complete", "core", and "enhanced" tumor components, respectively.

Acute and Chronic Plasma Metabalomic and Liver Transcriptomic Stress Effects in a Mouse Model with Features of Post-Traumatic Stress Disorder

DTIC Science & Technology

2015-01-28

trends is that at both 24 hr time points and the 1.5 wks time point, the HPA had already become desensitized , potentially involving attenuated release...points To explore metabolic alterations occurring at 24 hrs that potentially persisted out to 1.5 and 4 wks, we used random forests (RF) to classify...NR0B2, SLC27A3, SREBF1) Inhibition of cholesterol and lipid metabolism and transport; activation of Phase I metabolizing enzymes ->lipid and xenobiotic

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.

Text Categorization on Hadith Sahih Al-Bukhari using Random Forest

NASA Astrophysics Data System (ADS)

Fauzan Afianto, Muhammad; Adiwijaya; Al-Faraby, Said

2018-03-01

Al-Hadith is a collection of words, deeds, provisions, and approvals of Rasulullah Shallallahu Alaihi wa Salam that becomes the second fundamental laws of Islam after Al-Qur’an. As a fundamental of Islam, Muslims must learn, memorize, and practice Al-Qur’an and Al-Hadith. One of venerable Imam which was also the narrator of Al-Hadith is Imam Bukhari. He spent over 16 years to compile about 2602 Hadith (without repetition) and over 7000 Hadith with repetition. Automatic text categorization is a task of developing software tools that able to classify text of hypertext document under pre-defined categories or subject code[1]. The algorithm that would be used is Random Forest, which is a development from Decision Tree. In this final project research, the author decided to make a system that able to categorize text document that contains Hadith that narrated by Imam Bukhari under several categories such as suggestion, prohibition, and information. As for the evaluation method, K-fold cross validation with F1-Score will be used and the result is 90%.

Application of airborne hyperspectral remote sensing for the retrieval of forest inventory parameters

NASA Astrophysics Data System (ADS)

Dmitriev, Yegor V.; Kozoderov, Vladimir V.; Sokolov, Anton A.

2016-04-01

Collecting and updating forest inventory data play an important part in the forest management. The data can be obtained directly by using exact enough but low efficient ground based methods as well as from the remote sensing measurements. We present applications of airborne hyperspectral remote sensing for the retrieval of such important inventory parameters as the forest species and age composition. The hyperspectral images of the test region were obtained from the airplane equipped by the produced in Russia light-weight airborne video-spectrometer of visible and near infrared spectral range and high resolution photo-camera on the same gyro-stabilized platform. The quality of the thematic processing depends on many factors such as the atmospheric conditions, characteristics of measuring instruments, corrections and preprocessing methods, etc. An important role plays the construction of the classifier together with methods of the reduction of the feature space. The performance of different spectral classification methods is analyzed for the problem of hyperspectral remote sensing of soil and vegetation. For the reduction of the feature space we used the earlier proposed stable feature selection method. The results of the classification of hyperspectral airborne images by using the Multiclass Support Vector Machine method with Gaussian kernel and the parametric Bayesian classifier based on the Gaussian mixture model and their comparative analysis are demonstrated.

Discrimination of raw and processed Dipsacus asperoides by near infrared spectroscopy combined with least squares-support vector machine and random forests

NASA Astrophysics Data System (ADS)

Xin, Ni; Gu, Xiao-Feng; Wu, Hao; Hu, Yu-Zhu; Yang, Zhong-Lin

2012-04-01

Most herbal medicines could be processed to fulfill the different requirements of therapy. The purpose of this study was to discriminate between raw and processed Dipsacus asperoides, a common traditional Chinese medicine, based on their near infrared (NIR) spectra. Least squares-support vector machine (LS-SVM) and random forests (RF) were employed for full-spectrum classification. Three types of kernels, including linear kernel, polynomial kernel and radial basis function kernel (RBF), were checked for optimization of LS-SVM model. For comparison, a linear discriminant analysis (LDA) model was performed for classification, and the successive projections algorithm (SPA) was executed prior to building an LDA model to choose an appropriate subset of wavelengths. The three methods were applied to a dataset containing 40 raw herbs and 40 corresponding processed herbs. We ran 50 runs of 10-fold cross validation to evaluate the model's efficiency. The performance of the LS-SVM with RBF kernel (RBF LS-SVM) was better than the other two kernels. The RF, RBF LS-SVM and SPA-LDA successfully classified all test samples. The mean error rates for the 50 runs of 10-fold cross validation were 1.35% for RBF LS-SVM, 2.87% for RF, and 2.50% for SPA-LDA. The best classification results were obtained by using LS-SVM with RBF kernel, while RF was fast in the training and making predictions.

Bridging the gap between formal and experience-based knowledge for context-aware laparoscopy.

PubMed

Katić, Darko; Schuck, Jürgen; Wekerle, Anna-Laura; Kenngott, Hannes; Müller-Stich, Beat Peter; Dillmann, Rüdiger; Speidel, Stefanie

2016-06-01

Computer assistance is increasingly common in surgery. However, the amount of information is bound to overload processing abilities of surgeons. We propose methods to recognize the current phase of a surgery for context-aware information filtering. The purpose is to select the most suitable subset of information for surgical situations which require special assistance. We combine formal knowledge, represented by an ontology, and experience-based knowledge, represented by training samples, to recognize phases. For this purpose, we have developed two different methods. Firstly, we use formal knowledge about possible phase transitions to create a composition of random forests. Secondly, we propose a method based on cultural optimization to infer formal rules from experience to recognize phases. The proposed methods are compared with a purely formal knowledge-based approach using rules and a purely experience-based one using regular random forests. The comparative evaluation on laparoscopic pancreas resections and adrenalectomies employs a consistent set of quality criteria on clean and noisy input. The rule-based approaches proved best with noisefree data. The random forest-based ones were more robust in the presence of noise. Formal and experience-based knowledge can be successfully combined for robust phase recognition.

Inferring Planet Occurrence Rates With a Q1-Q17 Kepler Planet Candidate Catalog Produced by a Machine Learning Classifier

NASA Astrophysics Data System (ADS)

Catanzarite, Joseph; Jenkins, Jon Michael; McCauliff, Sean D.; Burke, Christopher; Bryson, Steve; Batalha, Natalie; Coughlin, Jeffrey; Rowe, Jason; mullally, fergal; thompson, susan; Seader, Shawn; Twicken, Joseph; Li, Jie; morris, robert; smith, jeffrey; haas, michael; christiansen, jessie; Clarke, Bruce

2015-08-01

NASA’s Kepler Space Telescope monitored the photometric variations of over 170,000 stars, at half-hour cadence, over its four-year prime mission. The Kepler pipeline calibrates the pixels of the target apertures for each star, produces light curves with simple aperture photometry, corrects for systematic error, and detects threshold-crossing events (TCEs) that may be due to transiting planets. The pipeline estimates planet parameters for all TCEs and computes diagnostics used by the Threshold Crossing Event Review Team (TCERT) to produce a catalog of objects that are deemed either likely transiting planet candidates or false positives.We created a training set from the Q1-Q12 and Q1-Q16 TCERT catalogs and an ensemble of synthetic transiting planets that were injected at the pixel level into all 17 quarters of data, and used it to train a random forest classifier. The classifier uniformly and consistently applies diagnostics developed by the Transiting Planet Search and Data Validation pipeline components and by TCERT to produce a robust catalog of planet candidates.The characteristics of the planet candidates detected by Kepler (planet radius and period) do not reflect the intrinsic planet population. Detection efficiency is a function of SNR, so the set of detected planet candidates is incomplete. Transit detection preferentially finds close-in planets with nearly edge-on orbits and misses planets whose orbital geometry precludes transits. Reliability of the planet candidates must also be considered, as they may be false positives. Errors in detected planet radius and in assumed star properties can also bias inference of intrinsic planet population characteristics.In this work we infer the intrinsic planet population, starting with the catalog of detected planet candidates produced by our random forest classifier, and accounting for detection biases and reliabilities as well as for radius errors in the detected population.Kepler was selected as the 10th mission of the Discovery Program. Funding for this mission is provided by NASA, Science Mission Directorate.

STACCATO: a novel solution to supernova photometric classification with biased training sets

NASA Astrophysics Data System (ADS)

Revsbech, E. A.; Trotta, R.; van Dyk, D. A.

2018-01-01

We present a new solution to the problem of classifying Type Ia supernovae from their light curves alone given a spectroscopically confirmed but biased training set, circumventing the need to obtain an observationally expensive unbiased training set. We use Gaussian processes (GPs) to model the supernovae's (SN's) light curves, and demonstrate that the choice of covariance function has only a small influence on the GPs ability to accurately classify SNe. We extend and improve the approach of Richards et al. - a diffusion map combined with a random forest classifier - to deal specifically with the case of biased training sets. We propose a novel method called Synthetically Augmented Light Curve Classification (STACCATO) that synthetically augments a biased training set by generating additional training data from the fitted GPs. Key to the success of the method is the partitioning of the observations into subgroups based on their propensity score of being included in the training set. Using simulated light curve data, we show that STACCATO increases performance, as measured by the area under the Receiver Operating Characteristic curve (AUC), from 0.93 to 0.96, close to the AUC of 0.977 obtained using the 'gold standard' of an unbiased training set and significantly improving on the previous best result of 0.88. STACCATO also increases the true positive rate for SNIa classification by up to a factor of 50 for high-redshift/low-brightness SNe.

Supervised fully polarimetric classification of the Black Forest test site: From MAESTROI to MAC Europe

NASA Technical Reports Server (NTRS)

Degrandi, G.; Lavalle, C.; Degroof, H.; Sieber, A.

1992-01-01

A study on the performance of a supervised fully polarimetric maximum likelihood classifier for synthetic aperture radar (SAR) data when applied to a specific classification context: forest classification based on age classes and in the presence of a sloping terrain is presented. For the experimental part, the polarimetric AIRSAR data at P, L, and C-band, acquired over the German Black Forest near Freiburg in the frame of the 1989 MAESTRO-1 campaign and the 1991 MAC Europe campaign was used, MAESTRO-1 with an ESA/JRC sponsored campaign, and MAC Europe (Multi-sensor Aircraft Campaign); in both cases the multi-frequency polarimetric JPL Airborne Synthetic Aperture Radar (AIRSAR) radar was flown over a number of European test sites. The study is structured as follows. At first, the general characteristics of the classifier and the dependencies from some parameters, like frequency bands, feature vector, calibration, using test areas lying on a flat terrain are investigated. Once it is determined the optimal conditions for the classifier performance, we then move on to the study of the slope effect. The bulk of this work is performed using the Maestrol data set. Next the classifier performance with the MAC Europe data is considered. The study is divided into two stages: first some of the tests done on the Maestro data are repeated, to highlight the improvements due to the new processing scheme that delivers 16 look data. Second we experiment with multi images classification with two goals: to assess the possibility of using a training set measured from one image to classify areas in different images; and to classify areas on critical slopes using different viewing angles. The main points of the study are listed and some of the results obtained so far are highlighted.

Stratifying to reduce bias caused by high nonresponse rates: A case study from New Mexico’s forest inventory

Treesearch

Sara A. Goeking; Paul L. Patterson

2013-01-01

The USDA Forest Service’s Forest Inventory and Analysis (FIA) Program applies specific sampling and analysis procedures to estimate a variety of forest attributes. FIA’s Interior West region uses post-stratification, where strata consist of forest/nonforest polygons based on MODIS imagery, and assumes that nonresponse plots are distributed at random across each stratum...

Forest cover loss and urban area expansion in the Conterminous Unites States in the first decade of the third millennium

NASA Astrophysics Data System (ADS)

Huo, L. Z.; Boschetti, L.

2016-12-01

Remote sensing has been successfully used for global mapping of changes in forest cover, but further analysis is needed to characterize those changes - and in particular to classify the total loss of forest loss (Gross Forest Cover Loss, GFCL) based on the cause (natural/human) and on the outcome of the change (regeneration to forest/transition to non-forest) (Kurtz et al., 2010). While natural forest disturbances (fires, insect outbreaks) and timber harvest generally involve a temporary change of land cover (vegetated to non-vegetated), they generally do not involve a change in land use, and it is expected that the forest cover loss is followed by recovery. Change of land use, such as the conversion of forest to agricultural or urban areas, is instead generally irreversible. The proper classification of forest cover loss is therefore necessary to properly model the long term effects of the disturbances on the carbon budget. The present study presents a spatial and temporal analysis of the forest cover loss due to urban expansion in the Conterminous United States. The Landsat-derived University of Maryland Global Forest Change product (Hansen et al, 2013) is used to identify all the areas of gross forest cover loss, which are subsequently classified into disturbance type (deforestation, stand-replacing natural disturbances, industrial forest clearcuts) using an object-oriented time series analysis (Huo and Boschetti, 2015). A further refinement of the classification is conducted to identify the areas of transition from forest land use to urban land use based on ancillary datasets such as the National Land Cover Database (Homer et al., 2015) and contextual image analysis techniques (analysis of object proximity, and detection of shapes). Results showed that over 4000 km2of forest were lost to urban area expansion in CONUS over the 2001 to 2010 period (1.8% of the gross forest cover loss). Most of the urban growth was concentrated in large urban areas: Atlanta, GA ranked first, followed by Houston, TX; Charlotte, NC; Jacksonville, FL; and Raleigh, NC. At the state level, the top 10 states with urban growth due to forest loss were GA, FL, TX, NC, SC, AL, LA, MS, VA and WA, which cumulatively accounted for 76 % of the total forest cover loss due to urban growth.

Application of random survival forests in understanding the determinants of under-five child mortality in Uganda in the presence of covariates that satisfy the proportional and non-proportional hazards assumption.

PubMed

Nasejje, Justine B; Mwambi, Henry

2017-09-07

Uganda just like any other Sub-Saharan African country, has a high under-five child mortality rate. To inform policy on intervention strategies, sound statistical methods are required to critically identify factors strongly associated with under-five child mortality rates. The Cox proportional hazards model has been a common choice in analysing data to understand factors strongly associated with high child mortality rates taking age as the time-to-event variable. However, due to its restrictive proportional hazards (PH) assumption, some covariates of interest which do not satisfy the assumption are often excluded in the analysis to avoid mis-specifying the model. Otherwise using covariates that clearly violate the assumption would mean invalid results. Survival trees and random survival forests are increasingly becoming popular in analysing survival data particularly in the case of large survey data and could be attractive alternatives to models with the restrictive PH assumption. In this article, we adopt random survival forests which have never been used in understanding factors affecting under-five child mortality rates in Uganda using Demographic and Health Survey data. Thus the first part of the analysis is based on the use of the classical Cox PH model and the second part of the analysis is based on the use of random survival forests in the presence of covariates that do not necessarily satisfy the PH assumption. Random survival forests and the Cox proportional hazards model agree that the sex of the household head, sex of the child, number of births in the past 1 year are strongly associated to under-five child mortality in Uganda given all the three covariates satisfy the PH assumption. Random survival forests further demonstrated that covariates that were originally excluded from the earlier analysis due to violation of the PH assumption were important in explaining under-five child mortality rates. These covariates include the number of children under the age of five in a household, number of births in the past 5 years, wealth index, total number of children ever born and the child's birth order. The results further indicated that the predictive performance for random survival forests built using covariates including those that violate the PH assumption was higher than that for random survival forests built using only covariates that satisfy the PH assumption. Random survival forests are appealing methods in analysing public health data to understand factors strongly associated with under-five child mortality rates especially in the presence of covariates that violate the proportional hazards assumption.

Ultrasonic Sensor Signals and Optimum Path Forest Classifier for the Microstructural Characterization of Thermally-Aged Inconel 625 Alloy

PubMed Central

de Albuquerque, Victor Hugo C.; Barbosa, Cleisson V.; Silva, Cleiton C.; Moura, Elineudo P.; Rebouças Filho, Pedro P.; Papa, João P.; Tavares, João Manuel R. S.

2015-01-01

Secondary phases, such as laves and carbides, are formed during the final solidification stages of nickel-based superalloy coatings deposited during the gas tungsten arc welding cold wire process. However, when aged at high temperatures, other phases can precipitate in the microstructure, like the γ” and δ phases. This work presents an evaluation of the powerful optimum path forest (OPF) classifier configured with six distance functions to classify background echo and backscattered ultrasonic signals from samples of the inconel 625 superalloy thermally aged at 650 and 950 °C for 10, 100 and 200 h. The background echo and backscattered ultrasonic signals were acquired using transducers with frequencies of 4 and 5 MHz. The potentiality of ultrasonic sensor signals combined with the OPF to characterize the microstructures of an inconel 625 thermally aged and in the as-welded condition were confirmed by the results. The experimental results revealed that the OPF classifier is sufficiently fast (classification total time of 0.316 ms) and accurate (accuracy of 88.75% and harmonic mean of 89.52) for the application proposed. PMID:26024416

Ultrasonic sensor signals and optimum path forest classifier for the microstructural characterization of thermally-aged inconel 625 alloy.

PubMed

de Albuquerque, Victor Hugo C; Barbosa, Cleisson V; Silva, Cleiton C; Moura, Elineudo P; Filho, Pedro P Rebouças; Papa, João P; Tavares, João Manuel R S

2015-05-27

Secondary phases, such as laves and carbides, are formed during the final solidification stages of nickel-based superalloy coatings deposited during the gas tungsten arc welding cold wire process. However, when aged at high temperatures, other phases can precipitate in the microstructure, like the γ'' and δ phases. This work presents an evaluation of the powerful optimum path forest (OPF) classifier configured with six distance functions to classify background echo and backscattered ultrasonic signals from samples of the inconel 625 superalloy thermally aged at 650 and 950 °C for 10, 100 and 200 h. The background echo and backscattered ultrasonic signals were acquired using transducers with frequencies of 4 and 5 MHz. The potentiality of ultrasonic sensor signals combined with the OPF to characterize the microstructures of an inconel 625 thermally aged and in the as-welded condition were confirmed by the results. The experimental results revealed that the OPF classifier is sufficiently fast (classification total time of 0.316 ms) and accurate (accuracy of 88.75%" and harmonic mean of 89.52) for the application proposed.

CNN-BLPred: a Convolutional neural network based predictor for β-Lactamases (BL) and their classes.

PubMed

White, Clarence; Ismail, Hamid D; Saigo, Hiroto; Kc, Dukka B

2017-12-28

The β-Lactamase (BL) enzyme family is an important class of enzymes that plays a key role in bacterial resistance to antibiotics. As the newly identified number of BL enzymes is increasing daily, it is imperative to develop a computational tool to classify the newly identified BL enzymes into one of its classes. There are two types of classification of BL enzymes: Molecular Classification and Functional Classification. Existing computational methods only address Molecular Classification and the performance of these existing methods is unsatisfactory. We addressed the unsatisfactory performance of the existing methods by implementing a Deep Learning approach called Convolutional Neural Network (CNN). We developed CNN-BLPred, an approach for the classification of BL proteins. The CNN-BLPred uses Gradient Boosted Feature Selection (GBFS) in order to select the ideal feature set for each BL classification. Based on the rigorous benchmarking of CCN-BLPred using both leave-one-out cross-validation and independent test sets, CCN-BLPred performed better than the other existing algorithms. Compared with other architectures of CNN, Recurrent Neural Network, and Random Forest, the simple CNN architecture with only one convolutional layer performs the best. After feature extraction, we were able to remove ~95% of the 10,912 features using Gradient Boosted Trees. During 10-fold cross validation, we increased the accuracy of the classic BL predictions by 7%. We also increased the accuracy of Class A, Class B, Class C, and Class D performance by an average of 25.64%. The independent test results followed a similar trend. We implemented a deep learning algorithm known as Convolutional Neural Network (CNN) to develop a classifier for BL classification. Combined with feature selection on an exhaustive feature set and using balancing method such as Random Oversampling (ROS), Random Undersampling (RUS) and Synthetic Minority Oversampling Technique (SMOTE), CNN-BLPred performs significantly better than existing algorithms for BL classification.

Research on electricity consumption forecast based on mutual information and random forests algorithm

NASA Astrophysics Data System (ADS)

Shi, Jing; Shi, Yunli; Tan, Jian; Zhu, Lei; Li, Hu

2018-02-01

Traditional power forecasting models cannot efficiently take various factors into account, neither to identify the relation factors. In this paper, the mutual information in information theory and the artificial intelligence random forests algorithm are introduced into the medium and long-term electricity demand prediction. Mutual information can identify the high relation factors based on the value of average mutual information between a variety of variables and electricity demand, different industries may be highly associated with different variables. The random forests algorithm was used for building the different industries forecasting models according to the different correlation factors. The data of electricity consumption in Jiangsu Province is taken as a practical example, and the above methods are compared with the methods without regard to mutual information and the industries. The simulation results show that the above method is scientific, effective, and can provide higher prediction accuracy.

Variable Selection for Road Segmentation in Aerial Images

NASA Astrophysics Data System (ADS)

Warnke, S.; Bulatov, D.

2017-05-01

For extraction of road pixels from combined image and elevation data, Wegner et al. (2015) proposed classification of superpixels into road and non-road, after which a refinement of the classification results using minimum cost paths and non-local optimization methods took place. We believed that the variable set used for classification was to a certain extent suboptimal, because many variables were redundant while several features known as useful in Photogrammetry and Remote Sensing are missed. This motivated us to implement a variable selection approach which builds a model for classification using portions of training data and subsets of features, evaluates this model, updates the feature set, and terminates when a stopping criterion is satisfied. The choice of classifier is flexible; however, we tested the approach with Logistic Regression and Random Forests, and taylored the evaluation module to the chosen classifier. To guarantee a fair comparison, we kept the segment-based approach and most of the variables from the related work, but we extended them by additional, mostly higher-level features. Applying these superior features, removing the redundant ones, as well as using more accurately acquired 3D data allowed to keep stable or even to reduce the misclassification error in a challenging dataset.

Forest disturbances, deforestation and timber harvest patterns in the Conterminous United States

NASA Astrophysics Data System (ADS)

Boschetti, L.; Huo, L. Z.

2016-12-01

Current estimates of carbon-equivalent emissions report the contribution of deforestation as 12% of total anthropogenic carbon emissions (van der Werf et al., 2009), but accurate monitoring of forest carbon balance should discriminate between land use change related to forest natural disturbances, forest management and deforestation. The total change in forest cover (Gross Forest Cover Loss, GFCL) needs to be characterized based on the cause (natural/human) and on the outcome of the change (regeneration to forest/transition to non-forest)(Kurtz et al, 2010). We developed a multitemporal, object-oriented methodology to classify GFCL as either (a) deforestation, (b) fire and insect disturbances (c) forest management practices. The Landsat-derived University of Maryland Global Forest Change product (Hansen, 2013) is used to identify all the areas forest cover loss: those areas are subsequently converted to objects, and used to extract temporal profiles of spectral reflectances and spectral indices from the Landsat WELD dataset. Finally, the temporal profiles and descriptive parameters of shapes, textures, and spatial relationships of the objects are used in a rule-based classifier to identify the type of disturbance. To pathfind a global disturbance type classification, the methods are demonstrated by wall-to-wall classification of the forest cover loss in the conterminous United States for the 2002-2011 period. The results show that deforestation accounts for a small percentage (approximately 2%) of the GFCL in the CONUS, and are in agreement with the known patterns of logging activity, fire and insect damage. The time series of timber harvest clearcut is also in agreement with the national timber extraction statistics, showing reduced harvesting following the 2008 economic crisis. The results also highlight the different management practices on private and public lands: 36% of the US forests are publicly owned (federal, state and local institutions) but account only for 12% of the clearcuts, whereas private lands (64% of the total) account for 88% of the clearcut area. Conversely, stand replacing fire and insect disturbances affect primarily public lands (85% versus 15% on private lands).

pSuc-Lys: Predict lysine succinylation sites in proteins with PseAAC and ensemble random forest approach.

PubMed

Jia, Jianhua; Liu, Zi; Xiao, Xuan; Liu, Bingxiang; Chou, Kuo-Chen

2016-04-07

Being one type of post-translational modifications (PTMs), protein lysine succinylation is important in regulating varieties of biological processes. It is also involved with some diseases, however. Consequently, from the angles of both basic research and drug development, we are facing a challenging problem: for an uncharacterized protein sequence having many Lys residues therein, which ones can be succinylated, and which ones cannot? To address this problem, we have developed a predictor called pSuc-Lys through (1) incorporating the sequence-coupled information into the general pseudo amino acid composition, (2) balancing out skewed training dataset by random sampling, and (3) constructing an ensemble predictor by fusing a series of individual random forest classifiers. Rigorous cross-validations indicated that it remarkably outperformed the existing methods. A user-friendly web-server for pSuc-Lys has been established at http://www.jci-bioinfo.cn/pSuc-Lys, by which users can easily obtain their desired results without the need to go through the complicated mathematical equations involved. It has not escaped our notice that the formulation and approach presented here can also be used to analyze many other problems in computational proteomics. Copyright © 2016 Elsevier Ltd. All rights reserved.

Discriminating the Mediterranean Pinus spp. using the land surface phenology extracted from the whole MODIS NDVI time series and machine learning algorithms

NASA Astrophysics Data System (ADS)

Rodriguez-Galiano, Victor; Aragones, David; Caparros-Santiago, Jose A.; Navarro-Cerrillo, Rafael M.

2017-10-01

Land surface phenology (LSP) can improve the characterisation of forest areas and their change processes. The aim of this work was: i) to characterise the temporal dynamics in Mediterranean Pinus forests, and ii) to evaluate the potential of LSP for species discrimination. The different experiments were based on 679 mono-specific plots for the 5 native species on the Iberian Peninsula: P. sylvestris, P. pinea, P. halepensis, P. nigra and P. pinaster. The entire MODIS NDVI time series (2000-2016) of the MOD13Q1 product was used to characterise phenology. The following phenological parameters were extracted: the start, end and median days of the season, and the length of the season in days, as well as the base value, maximum value, amplitude and integrated value. Multi-temporal metrics were calculated to synthesise the inter-annual variability of the phenological parameters. The species were discriminated by the application of Random Forest (RF) classifiers from different subsets of variables: model 1) NDVI-smoothed time series, model 2) multi-temporal metrics of the phenological parameters, and model 3) multi-temporal metrics and the auxiliary physical variables (altitude, slope, aspect and distance to the coastline). Model 3 was the best, with an overall accuracy of 82%, a kappa coefficient of 0.77 and whose most important variables were: elevation, coast distance, and the end and start days of the growing season. The species that presented the largest errors was P. nigra, (kappa= 0.45), having locations with a similar behaviour to P. sylvestris or P. pinaster.

Survival prediction of trauma patients: a study on US National Trauma Data Bank.

PubMed

Sefrioui, I; Amadini, R; Mauro, J; El Fallahi, A; Gabbrielli, M

2017-12-01

Exceptional circumstances like major incidents or natural disasters may cause a huge number of victims that might not be immediately and simultaneously saved. In these cases it is important to define priorities avoiding to waste time and resources for not savable victims. Trauma and Injury Severity Score (TRISS) methodology is the well-known and standard system usually used by practitioners to predict the survival probability of trauma patients. However, practitioners have noted that the accuracy of TRISS predictions is unacceptable especially for severely injured patients. Thus, alternative methods should be proposed. In this work we evaluate different approaches for predicting whether a patient will survive or not according to simple and easily measurable observations. We conducted a rigorous, comparative study based on the most important prediction techniques using real clinical data of the US National Trauma Data Bank. Empirical results show that well-known Machine Learning classifiers can outperform the TRISS methodology. Based on our findings, we can say that the best approach we evaluated is Random Forest: it has the best accuracy, the best area under the curve, and k-statistic, as well as the second-best sensitivity and specificity. It has also a good calibration curve. Furthermore, its performance monotonically increases as the dataset size grows, meaning that it can be very effective to exploit incoming knowledge. Considering the whole dataset, it is always better than TRISS. Finally, we implemented a new tool to compute the survival of victims. This will help medical practitioners to obtain a better accuracy than the TRISS tools. Random Forests may be a good candidate solution for improving the predictions on survival upon the standard TRISS methodology.

Automated diagnosis of myositis from muscle ultrasound: Exploring the use of machine learning and deep learning methods

PubMed Central

Burlina, Philippe; Billings, Seth; Joshi, Neil

2017-01-01

Objective To evaluate the use of ultrasound coupled with machine learning (ML) and deep learning (DL) techniques for automated or semi-automated classification of myositis. Methods Eighty subjects comprised of 19 with inclusion body myositis (IBM), 14 with polymyositis (PM), 14 with dermatomyositis (DM), and 33 normal (N) subjects were included in this study, where 3214 muscle ultrasound images of 7 muscles (observed bilaterally) were acquired. We considered three problems of classification including (A) normal vs. affected (DM, PM, IBM); (B) normal vs. IBM patients; and (C) IBM vs. other types of myositis (DM or PM). We studied the use of an automated DL method using deep convolutional neural networks (DL-DCNNs) for diagnostic classification and compared it with a semi-automated conventional ML method based on random forests (ML-RF) and “engineered” features. We used the known clinical diagnosis as the gold standard for evaluating performance of muscle classification. Results The performance of the DL-DCNN method resulted in accuracies ± standard deviation of 76.2% ± 3.1% for problem (A), 86.6% ± 2.4% for (B) and 74.8% ± 3.9% for (C), while the ML-RF method led to accuracies of 72.3% ± 3.3% for problem (A), 84.3% ± 2.3% for (B) and 68.9% ± 2.5% for (C). Conclusions This study demonstrates the application of machine learning methods for automatically or semi-automatically classifying inflammatory muscle disease using muscle ultrasound. Compared to the conventional random forest machine learning method used here, which has the drawback of requiring manual delineation of muscle/fat boundaries, DCNN-based classification by and large improved the accuracies in all classification problems while providing a fully automated approach to classification. PMID:28854220

Automated diagnosis of myositis from muscle ultrasound: Exploring the use of machine learning and deep learning methods.

PubMed

Burlina, Philippe; Billings, Seth; Joshi, Neil; Albayda, Jemima

2017-01-01

To evaluate the use of ultrasound coupled with machine learning (ML) and deep learning (DL) techniques for automated or semi-automated classification of myositis. Eighty subjects comprised of 19 with inclusion body myositis (IBM), 14 with polymyositis (PM), 14 with dermatomyositis (DM), and 33 normal (N) subjects were included in this study, where 3214 muscle ultrasound images of 7 muscles (observed bilaterally) were acquired. We considered three problems of classification including (A) normal vs. affected (DM, PM, IBM); (B) normal vs. IBM patients; and (C) IBM vs. other types of myositis (DM or PM). We studied the use of an automated DL method using deep convolutional neural networks (DL-DCNNs) for diagnostic classification and compared it with a semi-automated conventional ML method based on random forests (ML-RF) and "engineered" features. We used the known clinical diagnosis as the gold standard for evaluating performance of muscle classification. The performance of the DL-DCNN method resulted in accuracies ± standard deviation of 76.2% ± 3.1% for problem (A), 86.6% ± 2.4% for (B) and 74.8% ± 3.9% for (C), while the ML-RF method led to accuracies of 72.3% ± 3.3% for problem (A), 84.3% ± 2.3% for (B) and 68.9% ± 2.5% for (C). This study demonstrates the application of machine learning methods for automatically or semi-automatically classifying inflammatory muscle disease using muscle ultrasound. Compared to the conventional random forest machine learning method used here, which has the drawback of requiring manual delineation of muscle/fat boundaries, DCNN-based classification by and large improved the accuracies in all classification problems while providing a fully automated approach to classification.

Accelerating Convolutional Sparse Coding for Curvilinear Structures Segmentation by Refining SCIRD-TS Filter Banks.

PubMed

Annunziata, Roberto; Trucco, Emanuele

2016-11-01

Deep learning has shown great potential for curvilinear structure (e.g., retinal blood vessels and neurites) segmentation as demonstrated by a recent auto-context regression architecture based on filter banks learned by convolutional sparse coding. However, learning such filter banks is very time-consuming, thus limiting the amount of filters employed and the adaptation to other data sets (i.e., slow re-training). We address this limitation by proposing a novel acceleration strategy to speed-up convolutional sparse coding filter learning for curvilinear structure segmentation. Our approach is based on a novel initialisation strategy (warm start), and therefore it is different from recent methods improving the optimisation itself. Our warm-start strategy is based on carefully designed hand-crafted filters (SCIRD-TS), modelling appearance properties of curvilinear structures which are then refined by convolutional sparse coding. Experiments on four diverse data sets, including retinal blood vessels and neurites, suggest that the proposed method reduces significantly the time taken to learn convolutional filter banks (i.e., up to -82%) compared to conventional initialisation strategies. Remarkably, this speed-up does not worsen performance; in fact, filters learned with the proposed strategy often achieve a much lower reconstruction error and match or exceed the segmentation performance of random and DCT-based initialisation, when used as input to a random forest classifier.

Classifying northern forests using Thematic Mapper Simulator data

NASA Technical Reports Server (NTRS)

Nelson, R. F.; Latty, R. S.; Mott, G.

1984-01-01

Thematic Mapper Simulator data were collected over a 23,200 hectare forested area near Baxter State Park in north-central Maine. Photointerpreted ground reference information was used to drive a stratified random sampling procedure for waveband discriminant analyses and to generate training statistics and test pixel accuracies. Stepwise discriminant analyses indicated that the following bands best differentiated the thirteen level II - III cover types (in order of entry): near infrared (0.77 to 0.90 micron), blue (0.46 0.52 micron), first middle infrared (1.53 to 1.73 microns), second middle infrared (2.06 to 2.33 microsn), red (0.63 to 0.69 micron), thermal (10.32 to 12.33 microns). Classification accuracies peaked at 58 percent for thirteen level II-III land-cover classes and at 65 percent for ten level II classes.

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.

Applications of random forest feature selection for fine-scale genetic population assignment.

PubMed

Sylvester, Emma V A; Bentzen, Paul; Bradbury, Ian R; Clément, Marie; Pearce, Jon; Horne, John; Beiko, Robert G

2018-02-01

Genetic population assignment used to inform wildlife management and conservation efforts requires panels of highly informative genetic markers and sensitive assignment tests. We explored the utility of machine-learning algorithms (random forest, regularized random forest and guided regularized random forest) compared with F ST ranking for selection of single nucleotide polymorphisms (SNP) for fine-scale population assignment. We applied these methods to an unpublished SNP data set for Atlantic salmon ( Salmo salar ) and a published SNP data set for Alaskan Chinook salmon ( Oncorhynchus tshawytscha ). In each species, we identified the minimum panel size required to obtain a self-assignment accuracy of at least 90% using each method to create panels of 50-700 markers Panels of SNPs identified using random forest-based methods performed up to 7.8 and 11.2 percentage points better than F ST -selected panels of similar size for the Atlantic salmon and Chinook salmon data, respectively. Self-assignment accuracy ≥90% was obtained with panels of 670 and 384 SNPs for each data set, respectively, a level of accuracy never reached for these species using F ST -selected panels. Our results demonstrate a role for machine-learning approaches in marker selection across large genomic data sets to improve assignment for management and conservation of exploited populations.

A Minimum Spanning Forest Based Method for Noninvasive Cancer Detection with Hyperspectral Imaging

PubMed Central

Pike, Robert; Lu, Guolan; Wang, Dongsheng; Chen, Zhuo Georgia; Fei, Baowei

2016-01-01

Goal The purpose of this paper is to develop a classification method that combines both spectral and spatial information for distinguishing cancer from healthy tissue on hyperspectral images in an animal model. Methods An automated algorithm based on a minimum spanning forest (MSF) and optimal band selection has been proposed to classify healthy and cancerous tissue on hyperspectral images. A support vector machine (SVM) classifier is trained to create a pixel-wise classification probability map of cancerous and healthy tissue. This map is then used to identify markers that are used to compute mutual information for a range of bands in the hyperspectral image and thus select the optimal bands. An MSF is finally grown to segment the image using spatial and spectral information. Conclusion The MSF based method with automatically selected bands proved to be accurate in determining the tumor boundary on hyperspectral images. Significance Hyperspectral imaging combined with the proposed classification technique has the potential to provide a noninvasive tool for cancer detection. PMID:26285052

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.

Revealing the z ~ 2.5 Cosmic Web with 3D Lyα Forest Tomography: a Deformation Tensor Approach

NASA Astrophysics Data System (ADS)

Lee, Khee-Gan; White, Martin

2016-11-01

Studies of cosmological objects should take into account their positions within the cosmic web of large-scale structure. Unfortunately, the cosmic web has only been extensively mapped at low redshifts (z\\lt 1), using galaxy redshifts as tracers of the underlying density field. At z\\gt 1, the required galaxy densities are inaccessible for the foreseeable future, but 3D reconstructions of Lyα forest absorption in closely separated background QSOs and star-forming galaxies already offer a detailed window into z˜ 2-3 large-scale structure. We quantify the utility of such maps for studying the cosmic web by using realistic z = 2.5 Lyα forest simulations matched to observational properties of upcoming surveys. A deformation tensor-based analysis is used to classify voids, sheets, filaments, and nodes in the flux, which are compared to those determined from the underlying dark matter (DM) field. We find an extremely good correspondence, with 70% of the volume in the flux maps correctly classified relative to the DM web, and 99% classified to within one eigenvalue. This compares favorably to the performance of galaxy-based classifiers with even the highest galaxy densities from low-redshift surveys. We find that narrow survey geometries can degrade the recovery of the cosmic web unless the survey is ≳ 60 {h}-1 {Mpc} or ≳ 1 deg on the sky. We also examine halo abundances as a function of the cosmic web, and find a clear dependence as a function of flux overdensity, but little explicit dependence on the cosmic web. These methods will provide a new window on cosmological environments of galaxies at this very special time in galaxy formation, “high noon,” and on overall properties of cosmological structures at this epoch.

Decadal Trend in Agricultural Abandonment and Woodland Expansion in an Agro-Pastoral Transition Band in Northern China.

PubMed

Wang, Chao; Gao, Qiong; Wang, Xian; Yu, Mei

2015-01-01

Land use land cover (LULC) changes frequently in ecotones due to the large climate and soil gradients, and complex landscape composition and configuration. Accurate mapping of LULC changes in ecotones is of great importance for assessment of ecosystem functions/services and policy-decision support. Decadal or sub-decadal mapping of LULC provides scenarios for modeling biogeochemical processes and their feedbacks to climate, and evaluating effectiveness of land-use policies, e.g. forest conversion. However, it remains a great challenge to produce reliable LULC maps in moderate resolution and to evaluate their uncertainties over large areas with complex landscapes. In this study we developed a robust LULC classification system using multiple classifiers based on MODIS (Moderate Resolution Imaging Spectroradiometer) data and posterior data fusion. Not only does the system create LULC maps with high statistical accuracy, but also it provides pixel-level uncertainties that are essential for subsequent analyses and applications. We applied the classification system to the Agro-pasture transition band in northern China (APTBNC) to detect the decadal changes in LULC during 2003-2013 and evaluated the effectiveness of the implementation of major Key Forestry Programs (KFPs). In our study, the random forest (RF), support vector machine (SVM), and weighted k-nearest neighbors (WKNN) classifiers outperformed the artificial neural networks (ANN) and naive Bayes (NB) in terms of high classification accuracy and low sensitivity to training sample size. The Bayesian-average data fusion based on the results of RF, SVM, and WKNN achieved the 87.5% Kappa statistics, higher than any individual classifiers and the majority-vote integration. The pixel-level uncertainty map agreed with the traditional accuracy assessment. However, it conveys spatial variation of uncertainty. Specifically, it pinpoints the southwestern area of APTBNC has higher uncertainty than other part of the region, and the open shrubland is likely to be misclassified to the bare ground in some locations. Forests, closed shrublands, and grasslands in APTBNC expanded by 23%, 50%, and 9%, respectively, during 2003-2013. The expansion of these land cover types is compensated with the shrinkages in croplands (20%), bare ground (15%), and open shrublands (30%). The significant decline in agricultural lands is primarily attributed to the KFPs implemented in the end of last century and the nationwide urbanization in recent decade. The increased coverage of grass and woody plants would largely reduce soil erosion, improve mitigation of climate change, and enhance carbon sequestration in this region.

Mapping stand-age distribution of Russian forests from satellite data

NASA Astrophysics Data System (ADS)

Chen, D.; Loboda, T. V.; Hall, A.; Channan, S.; Weber, C. Y.

2013-12-01

Russian boreal forest is a critical component of the global boreal biome as approximately two thirds of the boreal forest is located in Russia. Numerous studies have shown that wildfire and logging have led to extensive modifications of forest cover in the region since 2000. Forest disturbance and subsequent regrowth influences carbon and energy budgets and, in turn, affect climate. Several global and regional satellite-based data products have been developed from coarse (>100m) and moderate (10-100m) resolution imagery to monitor forest cover change over the past decade, record of forest cover change pre-dating year 2000 is very fragmented. Although by using stacks of Landsat images, some information regarding the past disturbances can be obtained, the quantity and locations of such stacks with sufficient number of images are extremely limited, especially in Eastern Siberia. This paper describes a modified method which is built upon previous work to hindcast the disturbance history and map stand-age distribution in the Russian boreal forest. Utilizing data from both Landsat and the Moderate Resolution Imaging Spectroradiometer (MODIS), a wall-to-wall map indicating the estimated age of forest in the Russian boreal forest is created. Our previous work has shown that disturbances can be mapped successfully up to 30 years in the past as the spectral signature of regrowing forests is statistically significantly different from that of mature forests. The presented algorithm ingests 55 multi-temporal stacks of Landsat imagery available over Russian forest before 2001 and processes through a standardized and semi-automated approach to extract training and validation data samples. Landsat data, dating back to 1984, are used to generate maps of forest disturbance using temporal shifts in Disturbance Index through the multi-temporal stack of imagery in selected locations. These maps are then used as reference data to train a decision tree classifier on 50 MODIS-based indices. The resultant map provides an estimate of forest age based on the regrowth curves observed from Landsat imagery. The accuracy of the resultant map is assessed against three datasets: 1) subset of the disturbance maps developed within the algorithm, 2) independent disturbance maps created by the Northern Eurasia Land Dynamics Analysis (NELDA) project, and 3) field-based stand-age distribution from forestry inventory units. The current version of the product presents a considerable improvement on the previous version which used Landsat data samples at a set of randomly selected locations, resulting a strong bias of the training samples towards the Landsat-rich regions (e.g. European Russia) whereas regions such as Siberia were under-sampled. Aiming at improving accuracy, the current method significantly increases the number of training Landsat samples compared to the previous work. Aside from the previously used data, the current method uses all available Landsat data for the under-sampled regions in order to increase the representativeness of the total samples. The finial accuracy assessment is still ongoing, however, the initial results suggested an overall accuracy expressed in Kappa > 0.8. We plan to release both the training data and the final disturbance map of the Russian boreal forest to the public after the validation is completed.

Multi-Cohort Stand Structural Classification: Ground- and LiDAR-based Approaches for Boreal Mixedwood and Black Spruce Forest Types of Northeastern Ontario

NASA Astrophysics Data System (ADS)

Kuttner, Benjamin George

Natural fire return intervals are relatively long in eastern Canadian boreal forests and often allow for the development of stands with multiple, successive cohorts of trees. Multi-cohort forest management (MCM) provides a strategy to maintain such multi-cohort stands that focuses on three broad phases of increasingly complex, post-fire stand development, termed "cohorts", and recommends different silvicultural approaches be applied to emulate different cohort types. Previous research on structural cohort typing has relied upon primarily subjective classification methods; in this thesis, I develop more comprehensive and objective methods for three common boreal mixedwood and black spruce forest types in northeastern Ontario. Additionally, I examine relationships between cohort types and stand age, productivity, and disturbance history and the utility of airborne LiDAR to retrieve ground-based classifications and to extend structural cohort typing from plot- to stand-levels. In both mixedwood and black spruce forest types, stand age and age-related deadwood features varied systematically with cohort classes in support of an age-based interpretation of increasing cohort complexity. However, correlations of stand age with cohort classes were surprisingly weak. Differences in site productivity had a significant effect on the accrual of increasingly complex multi-cohort stand structure in both forest types, especially in black spruce stands. The effects of past harvesting in predictive models of class membership were only significant when considered in isolation of age. As an age-emulation strategy, the three cohort model appeared to be poorly suited to black spruce forests where the accrual of structural complexity appeared to be more a function of site productivity than age. Airborne LiDAR data appear to be particularly useful in recovering plot-based cohort types and extending them to the stand-level. The main gradients of structural variability detected using LiDAR were similar between boreal mixedwood and black spruce forest types; the best LiDAR-based models of cohort type relied upon combinations of tree size, size heterogeneity, and tree density related variables. The methods described here to measure, classify, and predict cohort-related structural complexity assist in translating the conceptual three cohort model to a more precise, measurement-based management system. In addition, the approaches presented here to measure and classify stand structural complexity promise to significantly enhance the detail of structural information in operational forest inventories in support of a wide array of forest management and conservation applications.

Integrating human and machine intelligence in galaxy morphology classification tasks

NASA Astrophysics Data System (ADS)

Beck, Melanie R.; Scarlata, Claudia; Fortson, Lucy F.; Lintott, Chris J.; Simmons, B. D.; Galloway, Melanie A.; Willett, Kyle W.; Dickinson, Hugh; Masters, Karen L.; Marshall, Philip J.; Wright, Darryl

2018-06-01

Quantifying galaxy morphology is a challenging yet scientifically rewarding task. As the scale of data continues to increase with upcoming surveys, traditional classification methods will struggle to handle the load. We present a solution through an integration of visual and automated classifications, preserving the best features of both human and machine. We demonstrate the effectiveness of such a system through a re-analysis of visual galaxy morphology classifications collected during the Galaxy Zoo 2 (GZ2) project. We reprocess the top-level question of the GZ2 decision tree with a Bayesian classification aggregation algorithm dubbed SWAP, originally developed for the Space Warps gravitational lens project. Through a simple binary classification scheme, we increase the classification rate nearly 5-fold classifying 226 124 galaxies in 92 d of GZ2 project time while reproducing labels derived from GZ2 classification data with 95.7 per cent accuracy. We next combine this with a Random Forest machine learning algorithm that learns on a suite of non-parametric morphology indicators widely used for automated morphologies. We develop a decision engine that delegates tasks between human and machine and demonstrate that the combined system provides at least a factor of 8 increase in the classification rate, classifying 210 803 galaxies in just 32 d of GZ2 project time with 93.1 per cent accuracy. As the Random Forest algorithm requires a minimal amount of computational cost, this result has important implications for galaxy morphology identification tasks in the era of Euclid and other large-scale surveys.

Prediction of soil attributes through interpolators in a deglaciated environment with complex landforms

NASA Astrophysics Data System (ADS)

Schünemann, Adriano Luis; Inácio Fernandes Filho, Elpídio; Rocha Francelino, Marcio; Rodrigues Santos, Gérson; Thomazini, Andre; Batista Pereira, Antônio; Gonçalves Reynaud Schaefer, Carlos Ernesto

2017-04-01

The knowledge of environmental variables values, in non-sampled sites from a minimum data set can be accessed through interpolation technique. Kriging and the classifier Random Forest algorithm are examples of predictors with this aim. The objective of this work was to compare methods of soil attributes spatialization in a recent deglaciated environment with complex landforms. Prediction of the selected soil attributes (potassium, calcium and magnesium) from ice-free areas were tested by using morphometric covariables, and geostatistical models without these covariables. For this, 106 soil samples were collected at 0-10 cm depth in Keller Peninsula, King George Island, Maritime Antarctica. Soil chemical analysis was performed by the gravimetric method, determining values of potassium, calcium and magnesium for each sampled point. Digital terrain models (DTMs) were obtained by using Terrestrial Laser Scanner. DTMs were generated from a cloud of points with spatial resolutions of 1, 5, 10, 20 and 30 m. Hence, 40 morphometric covariates were generated. Simple Kriging was performed using the R package software. The same data set coupled with morphometric covariates, was used to predict values of the studied attributes in non-sampled sites through Random Forest interpolator. Little differences were observed on the DTMs generated by Simple kriging and Random Forest interpolators. Also, DTMs with better spatial resolution did not improved the quality of soil attributes prediction. Results revealed that Simple Kriging can be used as interpolator when morphometric covariates are not available, with little impact regarding quality. It is necessary to go further in soil chemical attributes prediction techniques, especially in periglacial areas with complex landforms.

Nationwide classification of forest types of India using remote sensing and GIS.

PubMed

Reddy, C Sudhakar; Jha, C S; Diwakar, P G; Dadhwal, V K

2015-12-01

India, a mega-diverse country, possesses a wide range of climate and vegetation types along with a varied topography. The present study has classified forest types of India based on multi-season IRS Resourcesat-2 Advanced Wide Field Sensor (AWiFS) data. The study has characterized 29 land use/land cover classes including 14 forest types and seven scrub types. Hybrid classification approach has been used for the classification of forest types. The classification of vegetation has been carried out based on the ecological rule bases followed by Champion and Seth's (1968) scheme of forest types in India. The present classification scheme has been compared with the available global and national level land cover products. The natural vegetation cover was estimated to be 29.36% of total geographical area of India. The predominant forest types of India are tropical dry deciduous and tropical moist deciduous. Of the total forest cover, tropical dry deciduous forests occupy an area of 2,17,713 km(2) (34.80%) followed by 2,07,649 km(2) (33.19%) under tropical moist deciduous forests, 48,295 km(2) (7.72%) under tropical semi-evergreen forests and 47,192 km(2) (7.54%) under tropical wet evergreen forests. The study has brought out a comprehensive vegetation cover and forest type maps based on inputs critical in defining the various categories of vegetation and forest types. This spatially explicit database will be highly useful for the studies related to changes in various forest types, carbon stocks, climate-vegetation modeling and biogeochemical cycles.

Automated prediction of tissue outcome after acute ischemic stroke in computed tomography perfusion images

NASA Astrophysics Data System (ADS)

Vos, Pieter C.; Bennink, Edwin; de Jong, Hugo; Velthuis, Birgitta K.; Viergever, Max A.; Dankbaar, Jan Willem

2015-03-01

Assessment of the extent of cerebral damage on admission in patients with acute ischemic stroke could play an important role in treatment decision making. Computed tomography perfusion (CTP) imaging can be used to determine the extent of damage. However, clinical application is hindered by differences among vendors and used methodology. As a result, threshold based methods and visual assessment of CTP images has not yet shown to be useful in treatment decision making and predicting clinical outcome. Preliminary results in MR studies have shown the benefit of using supervised classifiers for predicting tissue outcome, but this has not been demonstrated for CTP. We present a novel method for the automatic prediction of tissue outcome by combining multi-parametric CTP images into a tissue outcome probability map. A supervised classification scheme was developed to extract absolute and relative perfusion values from processed CTP images that are summarized by a trained classifier into a likelihood of infarction. Training was performed using follow-up CT scans of 20 acute stroke patients with complete recanalization of the vessel that was occluded on admission. Infarcted regions were annotated by expert neuroradiologists. Multiple classifiers were evaluated in a leave-one-patient-out strategy for their discriminating performance using receiver operating characteristic (ROC) statistics. Results showed that a RandomForest classifier performed optimally with an area under the ROC of 0.90 for discriminating infarct tissue. The obtained results are an improvement over existing thresholding methods and are in line with results found in literature where MR perfusion was used.

Development of a computer aided diagnosis model for prostate cancer classification on multi-parametric MRI

NASA Astrophysics Data System (ADS)

Alfano, R.; Soetemans, D.; Bauman, G. S.; Gibson, E.; Gaed, M.; Moussa, M.; Gomez, J. A.; Chin, J. L.; Pautler, S.; Ward, A. D.

2018-02-01

Multi-parametric MRI (mp-MRI) is becoming a standard in contemporary prostate cancer screening and diagnosis, and has shown to aid physicians in cancer detection. It offers many advantages over traditional systematic biopsy, which has shown to have very high clinical false-negative rates of up to 23% at all stages of the disease. However beneficial, mp-MRI is relatively complex to interpret and suffers from inter-observer variability in lesion localization and grading. Computer-aided diagnosis (CAD) systems have been developed as a solution as they have the power to perform deterministic quantitative image analysis. We measured the accuracy of such a system validated using accurately co-registered whole-mount digitized histology. We trained a logistic linear classifier (LOGLC), support vector machine (SVC), k-nearest neighbour (KNN) and random forest classifier (RFC) in a four part ROI based experiment against: 1) cancer vs. non-cancer, 2) high-grade (Gleason score ≥4+3) vs. low-grade cancer (Gleason score <4+3), 3) high-grade vs. other tissue components and 4) high-grade vs. benign tissue by selecting the classifier with the highest AUC using 1-10 features from forward feature selection. The CAD model was able to classify malignant vs. benign tissue and detect high-grade cancer with high accuracy. Once fully validated, this work will form the basis for a tool that enhances the radiologist's ability to detect malignancies, potentially improving biopsy guidance, treatment selection, and focal therapy for prostate cancer patients, maximizing the potential for cure and increasing quality of life.

Prediction of Enzyme Mutant Activity Using Computational Mutagenesis and Incremental Transduction

PubMed Central

Basit, Nada; Wechsler, Harry

2011-01-01

Wet laboratory mutagenesis to determine enzyme activity changes is expensive and time consuming. This paper expands on standard one-shot learning by proposing an incremental transductive method (T2bRF) for the prediction of enzyme mutant activity during mutagenesis using Delaunay tessellation and 4-body statistical potentials for representation. Incremental learning is in tune with both eScience and actual experimentation, as it accounts for cumulative annotation effects of enzyme mutant activity over time. The experimental results reported, using cross-validation, show that overall the incremental transductive method proposed, using random forest as base classifier, yields better results compared to one-shot learning methods. T2bRF is shown to yield 90% on T4 and LAC (and 86% on HIV-1). This is significantly better than state-of-the-art competing methods, whose performance yield is at 80% or less using the same datasets. PMID:22007208

Superpixel-based structure classification for laparoscopic surgery

NASA Astrophysics Data System (ADS)

Bodenstedt, Sebastian; Görtler, Jochen; Wagner, Martin; Kenngott, Hannes; Müller-Stich, Beat Peter; Dillmann, Rüdiger; Speidel, Stefanie

2016-03-01

Minimally-invasive interventions offers multiple benefits for patients, but also entails drawbacks for the surgeon. The goal of context-aware assistance systems is to alleviate some of these difficulties. Localizing and identifying anatomical structures, maligned tissue and surgical instruments through endoscopic image analysis is paramount for an assistance system, making online measurements and augmented reality visualizations possible. Furthermore, such information can be used to assess the progress of an intervention, hereby allowing for a context-aware assistance. In this work, we present an approach for such an analysis. First, a given laparoscopic image is divided into groups of connected pixels, so-called superpixels, using the SEEDS algorithm. The content of a given superpixel is then described using information regarding its color and texture. Using a Random Forest classifier, we determine the class label of each superpixel. We evaluated our approach on a publicly available dataset for laparoscopic instrument detection and achieved a DICE score of 0.69.

Assessments of SENTINEL-2 Vegetation Red-Edge Spectral Bands for Improving Land Cover Classification

NASA Astrophysics Data System (ADS)

Qiu, S.; He, B.; Yin, C.; Liao, Z.

2017-09-01

The Multi Spectral Instrument (MSI) onboard Sentinel-2 can record the information in Vegetation Red-Edge (VRE) spectral domains. In this study, the performance of the VRE bands on improving land cover classification was evaluated based on a Sentinel-2A MSI image in East Texas, USA. Two classification scenarios were designed by excluding and including the VRE bands. A Random Forest (RF) classifier was used to generate land cover maps and evaluate the contributions of different spectral bands. The combination of VRE bands increased the overall classification accuracy by 1.40 %, which was statistically significant. Both confusion matrices and land cover maps indicated that the most beneficial increase was from vegetation-related land cover types, especially agriculture. Comparison of the relative importance of each band showed that the most beneficial VRE bands were Band 5 and Band 6. These results demonstrated the value of VRE bands for land cover classification.

Important LiDAR metrics for discriminating forest tree species in Central Europe

NASA Astrophysics Data System (ADS)

Shi, Yifang; Wang, Tiejun; Skidmore, Andrew K.; Heurich, Marco

2018-03-01

Numerous airborne LiDAR-derived metrics have been proposed for classifying tree species. Yet an in-depth ecological and biological understanding of the significance of these metrics for tree species mapping remains largely unexplored. In this paper, we evaluated the performance of 37 frequently used LiDAR metrics derived under leaf-on and leaf-off conditions, respectively, for discriminating six different tree species in a natural forest in Germany. We firstly assessed the correlation between these metrics. Then we applied a Random Forest algorithm to classify the tree species and evaluated the importance of the LiDAR metrics. Finally, we identified the most important LiDAR metrics and tested their robustness and transferability. Our results indicated that about 60% of LiDAR metrics were highly correlated to each other (|r| > 0.7). There was no statistically significant difference in tree species mapping accuracy between the use of leaf-on and leaf-off LiDAR metrics. However, combining leaf-on and leaf-off LiDAR metrics significantly increased the overall accuracy from 58.2% (leaf-on) and 62.0% (leaf-off) to 66.5% as well as the kappa coefficient from 0.47 (leaf-on) and 0.51 (leaf-off) to 0.58. Radiometric features, especially intensity related metrics, provided more consistent and significant contributions than geometric features for tree species discrimination. Specifically, the mean intensity of first-or-single returns as well as the mean value of echo width were identified as the most robust LiDAR metrics for tree species discrimination. These results indicate that metrics derived from airborne LiDAR data, especially radiometric metrics, can aid in discriminating tree species in a mixed temperate forest, and represent candidate metrics for tree species classification and monitoring in Central Europe.

Data-Driven Lead-Acid Battery Prognostics Using Random Survival Forests

DTIC Science & Technology

2014-10-02

Kogalur, Blackstone , & Lauer, 2008; Ishwaran & Kogalur, 2010). Random survival forest is a sur- vival analysis extension of Random Forests (Breiman, 2001...Statistics & probability letters, 80(13), 1056–1064. Ishwaran, H., Kogalur, U. B., Blackstone , E. H., & Lauer, M. S. (2008). Random survival forests. The

Amazon Rain Forest Classification Using J-ERS-1 SAR Data

NASA Technical Reports Server (NTRS)

Freeman, A.; Kramer, C.; Alves, M.; Chapman, B.

1994-01-01

The Amazon rain forest is a region of the earth that is undergoing rapid change. Man-made disturbance, such as clear cutting for agriculture or mining, is altering the rain forest ecosystem. For many parts of the rain forest, seasonal changes from the wet to the dry season are also significant. Changes in the seasonal cycle of flooding and draining can cause significant alterations in the forest ecosystem.Because much of the Amazon basin is regularly covered by thick clouds, optical and infrared coverage from the LANDSAT and SPOT satellites is sporadic. Imaging radar offers a much better potential for regular monitoring of changes in this region. In particular, the J-ERS-1 satellite carries an L-band HH SAR system, which via an on-board tape recorder, can collect data from almost anywhere on the globe at any time of year.In this paper, we show how J-ERS-1 radar images can be used to accurately classify different forest types (i.e., forest, hill forest, flooded forest), disturbed areas such as clear cuts and urban areas, and river courses in the Amazon basin. J-ERS-1 data has also shown significant differences between the dry and wet season, indicating a strong potential for monitoring seasonal change. The algorithm used to classify J-ERS-1 data is a standard maximum-likelihood classifier, using the radar image local mean and standard deviation of texture as input. Rivers and clear cuts are detected using edge detection and region-growing algorithms. Since this classifier is intended to operate successfully on data taken over the entire Amazon, several options are available to enable the user to modify the algorithm to suit a particular image.

Burnscar analysis using normalized burning ratio (NBR) index during 2015 forest fire at Merang-Kepahyang peat forest, South Sumatra, Indonesia

NASA Astrophysics Data System (ADS)

Saputra, Agus Dwi; Setiabudidaya, Dedi; Setyawan, Dwi; Khakim, M. Yusup Nur; Iskandar, Iskhaq

2017-07-01

Forest fire, classified as a natural hazard or human-induced hazard, has negative impacts on humans. These negative impacts are including economic loss, health problems, transportation disruption and land degradation or even biodiversity loss. During 2015, forest fire had occurred at the Merang-Kepahyang peat forest that has a total area of about 69.837,00 ha. In order to set a rehabilitation plan for recovering the impact of forest fire, information on the total burnscar area and severity level is required. In this study, the total burnscar area and severity level is evaluated using a calculation on the Normalized Burning Ratio (NBR) Index. The calculation is based on the Near Infra Red (NIR) and Short Wave Infra Red (SWIR) of the satellite imageries from LANDSAT. The images of pre-and post-fire are used to evaluate the severity level, which is defined as a difference in NBR Index of pre- and post-fire. It is found that about 42.906,00 ha of the total area of Merang-Kepahyang peat area have been fired in 2015. These burned area are classified into four categories, i.e., unburned, low, extreme and moderate extreme. By overlying the spatial map of burning level with other thematic maps, it is expected that strategy for rehabilitation plan can be well developed.

Floating Forests: Validation of a Citizen Science Effort to Answer Global Ecological Questions

NASA Astrophysics Data System (ADS)

Rosenthal, I.; Byrnes, J.; Cavanaugh, K. C.; Haupt, A. J.; Trouille, L.; Bell, T. W.; Rassweiler, A.; Pérez-Matus, A.; Assis, J.

2017-12-01

Researchers undertaking long term, large-scale ecological analyses face significant challenges for data collection and processing. Crowdsourcing via citizen science can provide an efficient method for analyzing large data sets. However, many scientists have raised questions about the quality of data collected by citizen scientists. Here we use Floating-Forests (http://floatingforests.org), a citizen science platform for creating a global time series of giant kelp abundance, to show that ensemble classifications of satellite data can ensure data quality. Citizen scientists view satellite images of coastlines and classify kelp forests by tracing all visible patches of kelp. Each image is classified by fifteen citizen scientists before being retired. To validate citizen science results, all fifteen classifications are converted to a raster and overlaid on a calibration dataset generated from previous studies. Results show that ensemble classifications from citizen scientists are consistently accurate when compared to calibration data. Given that all source images were acquired by Landsat satellites, we expect this consistency to hold across all regions. At present, we have over 6000 web-based citizen scientists' classifications of almost 2.5 million images of kelp forests in California and Tasmania. These results are not only useful for remote sensing of kelp forests, but also for a wide array of applications that combine citizen science with remote sensing.

Automatic migraine classification via feature selection committee and machine learning techniques over imaging and questionnaire data.

PubMed

Garcia-Chimeno, Yolanda; Garcia-Zapirain, Begonya; Gomez-Beldarrain, Marian; Fernandez-Ruanova, Begonya; Garcia-Monco, Juan Carlos

2017-04-13

Feature selection methods are commonly used to identify subsets of relevant features to facilitate the construction of models for classification, yet little is known about how feature selection methods perform in diffusion tensor images (DTIs). In this study, feature selection and machine learning classification methods were tested for the purpose of automating diagnosis of migraines using both DTIs and questionnaire answers related to emotion and cognition - factors that influence of pain perceptions. We select 52 adult subjects for the study divided into three groups: control group (15), subjects with sporadic migraine (19) and subjects with chronic migraine and medication overuse (18). These subjects underwent magnetic resonance with diffusion tensor to see white matter pathway integrity of the regions of interest involved in pain and emotion. The tests also gather data about pathology. The DTI images and test results were then introduced into feature selection algorithms (Gradient Tree Boosting, L1-based, Random Forest and Univariate) to reduce features of the first dataset and classification algorithms (SVM (Support Vector Machine), Boosting (Adaboost) and Naive Bayes) to perform a classification of migraine group. Moreover we implement a committee method to improve the classification accuracy based on feature selection algorithms. When classifying the migraine group, the greatest improvements in accuracy were made using the proposed committee-based feature selection method. Using this approach, the accuracy of classification into three types improved from 67 to 93% when using the Naive Bayes classifier, from 90 to 95% with the support vector machine classifier, 93 to 94% in boosting. The features that were determined to be most useful for classification included are related with the pain, analgesics and left uncinate brain (connected with the pain and emotions). The proposed feature selection committee method improved the performance of migraine diagnosis classifiers compared to individual feature selection methods, producing a robust system that achieved over 90% accuracy in all classifiers. The results suggest that the proposed methods can be used to support specialists in the classification of migraines in patients undergoing magnetic resonance imaging.

Application of remote sensing and GIS techniques for forest cover monitoring in the southern part of Laos

NASA Astrophysics Data System (ADS)

Keonuchan, Ammala; Liu, Yaolin

2008-12-01

Forest resource is the important material foundation of national sustainable development. And it need to master the status and change of forest resource timely for reasonable exploitation of forest and its renewal. Laos is located in the heart of the Indochinese peninsular, in southeast Asia, latitude 14° to 23 °north and longitude 100°to 108°east, covered a total 236, 800 square kilometers, and country of nearly 6 million people. The forest of Laos dropped from close to two-third in the 1970's to less than half by the 1990's. This deforestation has been attributed to two human activities : a traditional of shifting cultivation or slash and burn farming, and logging without reforestation. Remote sensing and GIS are the most modern technologies which have been widely used in the field of natural resource management and monitoring. These technologies provide very powerful tools to observe and collect information on natural resources and dynamic phenomenon on the earth surface, and ability to integrate different data and present data in different formats. In this study, using forest cover map and Landsat 7 ETM data, we analyze and compare forest cover change from 1997 to 2002. And the maximum likelihood method of supervised classification was used to classify the remote sensing data, we processed Spectral Enhancement, including Normalized Difference Vegetation Index (NDVI) ,and re-classify data again base on Principle Components Analysis (PCA) and NDVI.

Detection of dead standing Eucalyptus camaldulensis without tree delineation for managing biodiversity in native Australian forest

NASA Astrophysics Data System (ADS)

Miltiadou, Milto; Campbell, Neil D. F.; Gonzalez Aracil, Susana; Brown, Tony; Grant, Michael G.

2018-05-01

In Australia, many birds and arboreal animals use hollows for shelters, but studies predict shortage of hollows in near future. Aged dead trees are more likely to contain hollows and therefore automated detection of them plays a substantial role in preserving biodiversity and consequently maintaining a resilient ecosystem. For this purpose full-waveform LiDAR data were acquired from a native Eucalypt forest in Southern Australia. The structure of the forest significantly varies in terms of tree density, age and height. Additionally, Eucalyptus camaldulensis have multiple trunk splits making tree delineation very challenging. For that reason, this paper investigates automated detection of dead standing Eucalyptus camaldulensis without tree delineation. It also presents the new feature of the open source software DASOS, which extracts features for 3D object detection in voxelised FW LiDAR. A random forest classifier, a weighted-distance KNN algorithm and a seed growth algorithm are used to create a 2D probabilistic field and to then predict potential positions of dead trees. It is shown that tree health assessment is possible without tree delineation but since it is a new research directions there are many improvements to be made.

Forest Cover Mapping in Iskandar Malaysia Using Satellite Data

NASA Astrophysics Data System (ADS)

Kanniah, K. D.; Mohd Najib, N. E.; Vu, T. T.

2016-09-01

Malaysia is the third largest country in the world that had lost forest cover. Therefore, timely information on forest cover is required to help the government to ensure that the remaining forest resources are managed in a sustainable manner. This study aims to map and detect changes of forest cover (deforestation and disturbance) in Iskandar Malaysia region in the south of Peninsular Malaysia between years 1990 and 2010 using Landsat satellite images. The Carnegie Landsat Analysis System-Lite (CLASlite) programme was used to classify forest cover using Landsat images. This software is able to mask out clouds, cloud shadows, terrain shadows, and water bodies and atmospherically correct the images using 6S radiative transfer model. An Automated Monte Carlo Unmixing technique embedded in CLASlite was used to unmix each Landsat pixel into fractions of photosynthetic vegetation (PV), non photosynthetic vegetation (NPV) and soil surface (S). Forest and non-forest areas were produced from the fractional cover images using appropriate threshold values of PV, NPV and S. CLASlite software was found to be able to classify forest cover in Iskandar Malaysia with only a difference between 14% (1990) and 5% (2010) compared to the forest land use map produced by the Department of Agriculture, Malaysia. Nevertheless, the CLASlite automated software used in this study was found not to exclude other vegetation types especially rubber and oil palm that has similar reflectance to forest. Currently rubber and oil palm were discriminated from forest manually using land use maps. Therefore, CLASlite algorithm needs further adjustment to exclude these vegetation and classify only forest cover.

Foliar and woody materials discriminated using terrestrial LiDAR in a mixed natural forest

NASA Astrophysics Data System (ADS)

Zhu, Xi; Skidmore, Andrew K.; Darvishzadeh, Roshanak; Niemann, K. Olaf; Liu, Jing; Shi, Yifang; Wang, Tiejun

2018-02-01

Separation of foliar and woody materials using remotely sensed data is crucial for the accurate estimation of leaf area index (LAI) and woody biomass across forest stands. In this paper, we present a new method to accurately separate foliar and woody materials using terrestrial LiDAR point clouds obtained from ten test sites in a mixed forest in Bavarian Forest National Park, Germany. Firstly, we applied and compared an adaptive radius near-neighbor search algorithm with a fixed radius near-neighbor search method in order to obtain both radiometric and geometric features derived from terrestrial LiDAR point clouds. Secondly, we used a random forest machine learning algorithm to classify foliar and woody materials and examined the impact of understory and slope on the classification accuracy. An average overall accuracy of 84.4% (Kappa = 0.75) was achieved across all experimental plots. The adaptive radius near-neighbor search method outperformed the fixed radius near-neighbor search method. The classification accuracy was significantly higher when the combination of both radiometric and geometric features was utilized. The analysis showed that increasing slope and understory coverage had a significant negative effect on the overall classification accuracy. Our results suggest that the utilization of the adaptive radius near-neighbor search method coupling both radiometric and geometric features has the potential to accurately discriminate foliar and woody materials from terrestrial LiDAR data in a mixed natural forest.

Screening large-scale association study data: exploiting interactions using random forests.

PubMed

Lunetta, Kathryn L; Hayward, L Brooke; Segal, Jonathan; Van Eerdewegh, Paul

2004-12-10

Genome-wide association studies for complex diseases will produce genotypes on hundreds of thousands of single nucleotide polymorphisms (SNPs). A logical first approach to dealing with massive numbers of SNPs is to use some test to screen the SNPs, retaining only those that meet some criterion for further study. For example, SNPs can be ranked by p-value, and those with the lowest p-values retained. When SNPs have large interaction effects but small marginal effects in a population, they are unlikely to be retained when univariate tests are used for screening. However, model-based screens that pre-specify interactions are impractical for data sets with thousands of SNPs. Random forest analysis is an alternative method that produces a single measure of importance for each predictor variable that takes into account interactions among variables without requiring model specification. Interactions increase the importance for the individual interacting variables, making them more likely to be given high importance relative to other variables. We test the performance of random forests as a screening procedure to identify small numbers of risk-associated SNPs from among large numbers of unassociated SNPs using complex disease models with up to 32 loci, incorporating both genetic heterogeneity and multi-locus interaction. Keeping other factors constant, if risk SNPs interact, the random forest importance measure significantly outperforms the Fisher Exact test as a screening tool. As the number of interacting SNPs increases, the improvement in performance of random forest analysis relative to Fisher Exact test for screening also increases. Random forests perform similarly to the univariate Fisher Exact test as a screening tool when SNPs in the analysis do not interact. In the context of large-scale genetic association studies where unknown interactions exist among true risk-associated SNPs or SNPs and environmental covariates, screening SNPs using random forest analyses can significantly reduce the number of SNPs that need to be retained for further study compared to standard univariate screening methods.

Random forests on Hadoop for genome-wide association studies of multivariate neuroimaging phenotypes

PubMed Central

2013-01-01

Motivation Multivariate quantitative traits arise naturally in recent neuroimaging genetics studies, in which both structural and functional variability of the human brain is measured non-invasively through techniques such as magnetic resonance imaging (MRI). There is growing interest in detecting genetic variants associated with such multivariate traits, especially in genome-wide studies. Random forests (RFs) classifiers, which are ensembles of decision trees, are amongst the best performing machine learning algorithms and have been successfully employed for the prioritisation of genetic variants in case-control studies. RFs can also be applied to produce gene rankings in association studies with multivariate quantitative traits, and to estimate genetic similarities measures that are predictive of the trait. However, in studies involving hundreds of thousands of SNPs and high-dimensional traits, a very large ensemble of trees must be inferred from the data in order to obtain reliable rankings, which makes the application of these algorithms computationally prohibitive. Results We have developed a parallel version of the RF algorithm for regression and genetic similarity learning tasks in large-scale population genetic association studies involving multivariate traits, called PaRFR (Parallel Random Forest Regression). Our implementation takes advantage of the MapReduce programming model and is deployed on Hadoop, an open-source software framework that supports data-intensive distributed applications. Notable speed-ups are obtained by introducing a distance-based criterion for node splitting in the tree estimation process. PaRFR has been applied to a genome-wide association study on Alzheimer's disease (AD) in which the quantitative trait consists of a high-dimensional neuroimaging phenotype describing longitudinal changes in the human brain structure. PaRFR provides a ranking of SNPs associated to this trait, and produces pair-wise measures of genetic proximity that can be directly compared to pair-wise measures of phenotypic proximity. Several known AD-related variants have been identified, including APOE4 and TOMM40. We also present experimental evidence supporting the hypothesis of a linear relationship between the number of top-ranked mutated states, or frequent mutation patterns, and an indicator of disease severity. Availability The Java codes are freely available at http://www2.imperial.ac.uk/~gmontana. PMID:24564704

Random forests on Hadoop for genome-wide association studies of multivariate neuroimaging phenotypes.

PubMed

Wang, Yue; Goh, Wilson; Wong, Limsoon; Montana, Giovanni

2013-01-01

Multivariate quantitative traits arise naturally in recent neuroimaging genetics studies, in which both structural and functional variability of the human brain is measured non-invasively through techniques such as magnetic resonance imaging (MRI). There is growing interest in detecting genetic variants associated with such multivariate traits, especially in genome-wide studies. Random forests (RFs) classifiers, which are ensembles of decision trees, are amongst the best performing machine learning algorithms and have been successfully employed for the prioritisation of genetic variants in case-control studies. RFs can also be applied to produce gene rankings in association studies with multivariate quantitative traits, and to estimate genetic similarities measures that are predictive of the trait. However, in studies involving hundreds of thousands of SNPs and high-dimensional traits, a very large ensemble of trees must be inferred from the data in order to obtain reliable rankings, which makes the application of these algorithms computationally prohibitive. We have developed a parallel version of the RF algorithm for regression and genetic similarity learning tasks in large-scale population genetic association studies involving multivariate traits, called PaRFR (Parallel Random Forest Regression). Our implementation takes advantage of the MapReduce programming model and is deployed on Hadoop, an open-source software framework that supports data-intensive distributed applications. Notable speed-ups are obtained by introducing a distance-based criterion for node splitting in the tree estimation process. PaRFR has been applied to a genome-wide association study on Alzheimer's disease (AD) in which the quantitative trait consists of a high-dimensional neuroimaging phenotype describing longitudinal changes in the human brain structure. PaRFR provides a ranking of SNPs associated to this trait, and produces pair-wise measures of genetic proximity that can be directly compared to pair-wise measures of phenotypic proximity. Several known AD-related variants have been identified, including APOE4 and TOMM40. We also present experimental evidence supporting the hypothesis of a linear relationship between the number of top-ranked mutated states, or frequent mutation patterns, and an indicator of disease severity. The Java codes are freely available at http://www2.imperial.ac.uk/~gmontana.

Multidecadal Rates of Disturbance- and Climate Change-Induced Land Cover Change in Arctic and Boreal Ecosystems over Western Canada and Alaska Inferred from Dense Landsat Time Series

NASA Astrophysics Data System (ADS)

Wang, J.; Sulla-menashe, D. J.; Woodcock, C. E.; Sonnentag, O.; Friedl, M. A.

2017-12-01

Rapid climate change in arctic and boreal ecosystems is driving changes to land cover composition, including woody expansion in the arctic tundra, successional shifts following boreal fires, and thaw-induced wetland expansion and forest collapse along the southern limit of permafrost. The impacts of these land cover transformations on the physical climate and the carbon cycle are increasingly well-documented from field and model studies, but there have been few attempts to empirically estimate rates of land cover change at decadal time scale and continental spatial scale. Previous studies have used too coarse spatial resolution or have been too limited in temporal range to enable broad multi-decadal assessment of land cover change. As part of NASA's Arctic Boreal Vulnerability Experiment (ABoVE), we are using dense time series of Landsat remote sensing data to map disturbances and classify land cover types across the ABoVE extended domain (spanning western Canada and Alaska) over the last three decades (1982-2014) at 30 m resolution. We utilize regionally-complete and repeated acquisition high-resolution (<2 m) DigitalGlobe imagery to generate training data from across the region that follows a nested, hierarchical classification scheme encompassing plant functional type and cover density, understory type, wetland status, and land use. Additionally, we crosswalk plot-level field data into our scheme for additional high quality training sites. We use the Continuous Change Detection and Classification algorithm to estimate land cover change dates and temporal-spectral features in the Landsat data. These features are used to train random forest classification models and map land cover and analyze land cover change processes, focusing primarily on tundra "shrubification", post-fire succession, and boreal wetland expansion. We will analyze the high resolution data based on stratified random sampling of our change maps to validate and assess the accuracy of our model predictions. In this paper, we present initial results from this effort, including sub-regional analyses focused on several key areas, such as the Taiga Plains and the Southern Arctic ecozones, to calibrate our random forest models and assess results.

An exploratory study of a text classification framework for Internet-based surveillance of emerging epidemics

PubMed Central

Torii, Manabu; Yin, Lanlan; Nguyen, Thang; Mazumdar, Chand T.; Liu, Hongfang; Hartley, David M.; Nelson, Noele P.

2014-01-01

Purpose Early detection of infectious disease outbreaks is crucial to protecting the public health of a society. Online news articles provide timely information on disease outbreaks worldwide. In this study, we investigated automated detection of articles relevant to disease outbreaks using machine learning classifiers. In a real-life setting, it is expensive to prepare a training data set for classifiers, which usually consists of manually labeled relevant and irrelevant articles. To mitigate this challenge, we examined the use of randomly sampled unlabeled articles as well as labeled relevant articles. Methods Naïve Bayes and Support Vector Machine (SVM) classifiers were trained on 149 relevant and 149 or more randomly sampled unlabeled articles. Diverse classifiers were trained by varying the number of sampled unlabeled articles and also the number of word features. The trained classifiers were applied to 15 thousand articles published over 15 days. Top-ranked articles from each classifier were pooled and the resulting set of 1337 articles was reviewed by an expert analyst to evaluate the classifiers. Results Daily averages of areas under ROC curves (AUCs) over the 15-day evaluation period were 0.841 and 0.836, respectively, for the naïve Bayes and SVM classifier. We referenced a database of disease outbreak reports to confirm that this evaluation data set resulted from the pooling method indeed covered incidents recorded in the database during the evaluation period. Conclusions The proposed text classification framework utilizing randomly sampled unlabeled articles can facilitate a cost-effective approach to training machine learning classifiers in a real-life Internet-based biosurveillance project. We plan to examine this framework further using larger data sets and using articles in non-English languages. PMID:21134784

Models for H₃ receptor antagonist activity of sulfonylurea derivatives.

PubMed

Khatri, Naveen; Madan, A K

2014-03-01

The histamine H₃ receptor has been perceived as an auspicious target for the treatment of various central and peripheral nervous system diseases. In present study, a wide variety of 60 2D and 3D molecular descriptors (MDs) were successfully utilized for the development of models for the prediction of antagonist activity of sulfonylurea derivatives for histamine H₃ receptors. Models were developed through decision tree (DT), random forest (RF) and moving average analysis (MAA). Dragon software version 6.0.28 was employed for calculation of values of diverse MDs of each analogue involved in the data set. The DT classified and correctly predicted the input data with an impressive non-error rate of 94% in the training set and 82.5% during cross validation. RF correctly classified the analogues into active and inactive with a non-error rate of 79.3%. The MAA based models predicted the antagonist histamine H₃ receptor activity with non-error rate up to 90%. Active ranges of the proposed MAA based models not only exhibited high potency but also showed improved safety as indicated by relatively high values of selectivity index. The statistical significance of the models was assessed through sensitivity, specificity, non-error rate, Matthew's correlation coefficient and intercorrelation analysis. Proposed models offer vast potential for providing lead structures for development of potent but safe H₃ receptor antagonist sulfonylurea derivatives. Copyright © 2013 Elsevier Inc. All rights reserved.

Cascade Classification with Adaptive Feature Extraction for Arrhythmia Detection.

PubMed

Park, Juyoung; Kang, Mingon; Gao, Jean; Kim, Younghoon; Kang, Kyungtae

2017-01-01

Detecting arrhythmia from ECG data is now feasible on mobile devices, but in this environment it is necessary to trade computational efficiency against accuracy. We propose an adaptive strategy for feature extraction that only considers normalized beat morphology features when running in a resource-constrained environment; but in a high-performance environment it takes account of a wider range of ECG features. This process is augmented by a cascaded random forest classifier. Experiments on data from the MIT-BIH Arrhythmia Database showed classification accuracies from 96.59% to 98.51%, which are comparable to state-of-the art methods.

Random Forest Application for NEXRAD Radar Data Quality Control

NASA Astrophysics Data System (ADS)

Keem, M.; Seo, B. C.; Krajewski, W. F.

2017-12-01

Identification and elimination of non-meteorological radar echoes (e.g., returns from ground, wind turbines, and biological targets) are the basic data quality control steps before radar data use in quantitative applications (e.g., precipitation estimation). Although WSR-88Ds' recent upgrade to dual-polarization has enhanced this quality control and echo classification, there are still challenges to detect some non-meteorological echoes that show precipitation-like characteristics (e.g., wind turbine or anomalous propagation clutter embedded in rain). With this in mind, a new quality control method using Random Forest is proposed in this study. This classification algorithm is known to produce reliable results with less uncertainty. The method introduces randomness into sampling and feature selections and integrates consequent multiple decision trees. The multidimensional structure of the trees can characterize the statistical interactions of involved multiple features in complex situations. The authors explore the performance of Random Forest method for NEXRAD radar data quality control. Training datasets are selected using several clear cases of precipitation and non-precipitation (but with some non-meteorological echoes). The model is structured using available candidate features (from the NEXRAD data) such as horizontal reflectivity, differential reflectivity, differential phase shift, copolar correlation coefficient, and their horizontal textures (e.g., local standard deviation). The influence of each feature on classification results are quantified by variable importance measures that are automatically estimated by the Random Forest algorithm. Therefore, the number and types of features in the final forest can be examined based on the classification accuracy. The authors demonstrate the capability of the proposed approach using several cases ranging from distinct to complex rain/no-rain events and compare the performance with the existing algorithms (e.g., MRMS). They also discuss operational feasibility based on the observed strength and weakness of the method.

BCDForest: a boosting cascade deep forest model towards the classification of cancer subtypes based on gene expression data.

PubMed

Guo, Yang; Liu, Shuhui; Li, Zhanhuai; Shang, Xuequn

2018-04-11

The classification of cancer subtypes is of great importance to cancer disease diagnosis and therapy. Many supervised learning approaches have been applied to cancer subtype classification in the past few years, especially of deep learning based approaches. Recently, the deep forest model has been proposed as an alternative of deep neural networks to learn hyper-representations by using cascade ensemble decision trees. It has been proved that the deep forest model has competitive or even better performance than deep neural networks in some extent. However, the standard deep forest model may face overfitting and ensemble diversity challenges when dealing with small sample size and high-dimensional biology data. In this paper, we propose a deep learning model, so-called BCDForest, to address cancer subtype classification on small-scale biology datasets, which can be viewed as a modification of the standard deep forest model. The BCDForest distinguishes from the standard deep forest model with the following two main contributions: First, a named multi-class-grained scanning method is proposed to train multiple binary classifiers to encourage diversity of ensemble. Meanwhile, the fitting quality of each classifier is considered in representation learning. Second, we propose a boosting strategy to emphasize more important features in cascade forests, thus to propagate the benefits of discriminative features among cascade layers to improve the classification performance. Systematic comparison experiments on both microarray and RNA-Seq gene expression datasets demonstrate that our method consistently outperforms the state-of-the-art methods in application of cancer subtype classification. The multi-class-grained scanning and boosting strategy in our model provide an effective solution to ease the overfitting challenge and improve the robustness of deep forest model working on small-scale data. Our model provides a useful approach to the classification of cancer subtypes by using deep learning on high-dimensional and small-scale biology data.

Quantitative CT analysis for the preoperative prediction of pathologic grade in pancreatic neuroendocrine tumors

NASA Astrophysics Data System (ADS)

Chakraborty, Jayasree; Pulvirenti, Alessandra; Yamashita, Rikiya; Midya, Abhishek; Gönen, Mithat; Klimstra, David S.; Reidy, Diane L.; Allen, Peter J.; Do, Richard K. G.; Simpson, Amber L.

2018-02-01

Pancreatic neuroendocrine tumors (PanNETs) account for approximately 5% of all pancreatic tumors, affecting one individual per million each year.1 PanNETs are difficult to treat due to biological variability from benign to highly malignant, indolent to very aggressive. The World Health Organization classifies PanNETs into three categories based on cell proliferative rate, usually detected using the Ki67 index and cell morphology: low-grade (G1), intermediate-grade (G2) and high-grade (G3) tumors. Knowledge of grade prior to treatment would select patients for optimal therapy: G1/G2 tumors respond well to somatostatin analogs and targeted or cytotoxic drugs whereas G3 tumors would be targeted with platinum or alkylating agents.2, 3 Grade assessment is based on the pathologic examination of the surgical specimen, biopsy or ne-needle aspiration; however, heterogeneity in the proliferative index can lead to sampling errors.4 Based on studies relating qualitatively assessed shape and enhancement characteristics on CT imaging to tumor grade in PanNET,5 we propose objective classification of PanNET grade with quantitative analysis of CT images. Fifty-five patients were included in our retrospective analysis. A pathologist graded the tumors. Texture and shape-based features were extracted from CT. Random forest and naive Bayes classifiers were compared for the classification of G1/G2 and G3 PanNETs. The best area under the receiver operating characteristic curve (AUC) of 0:74 and accuracy of 71:64% was achieved with texture features. The shape-based features achieved an AUC of 0:70 and accuracy of 78:73%.

The Efficiency of Random Forest Method for Shoreline Extraction from LANDSAT-8 and GOKTURK-2 Imageries

NASA Astrophysics Data System (ADS)

Bayram, B.; Erdem, F.; Akpinar, B.; Ince, A. K.; Bozkurt, S.; Catal Reis, H.; Seker, D. Z.

2017-11-01

Coastal monitoring plays a vital role in environmental planning and hazard management related issues. Since shorelines are fundamental data for environment management, disaster management, coastal erosion studies, modelling of sediment transport and coastal morphodynamics, various techniques have been developed to extract shorelines. Random Forest is one of these techniques which is used in this study for shoreline extraction.. This algorithm is a machine learning method based on decision trees. Decision trees analyse classes of training data creates rules for classification. In this study, Terkos region has been chosen for the proposed method within the scope of "TUBITAK Project (Project No: 115Y718) titled "Integration of Unmanned Aerial Vehicles for Sustainable Coastal Zone Monitoring Model - Three-Dimensional Automatic Coastline Extraction and Analysis: Istanbul-Terkos Example". Random Forest algorithm has been implemented to extract the shoreline of the Black Sea where near the lake from LANDSAT-8 and GOKTURK-2 satellite imageries taken in 2015. The MATLAB environment was used for classification. To obtain land and water-body classes, the Random Forest method has been applied to NIR bands of LANDSAT-8 (5th band) and GOKTURK-2 (4th band) imageries. Each image has been digitized manually and shorelines obtained for accuracy assessment. According to accuracy assessment results, Random Forest method is efficient for both medium and high resolution images for shoreline extraction studies.

New York Forests, 2012

Treesearch

Richard H. Widmann; Sloane Crawford; Cassandra M. Kurtz; Mark D. Nelson; Patrick D. Miles; Randall S. Morin; Rachel. Riemann

2015-01-01

This report summarizes the second annual inventory of New York's forests, conducted in 2008-2012. New York's forests cover 19.0 million acres; 15.9 million acres are classified as timberland and 3.1 million acres as reserved and other forest land. Forest land is dominated by the maple/beech/birch forest-type group that occupies more than half of the forest...

Multi-categorical deep learning neural network to classify retinal images: A pilot study employing small database

PubMed Central

Seo, Jeong Gi; Kwak, Jiyong; Um, Terry Taewoong; Rim, Tyler Hyungtaek

2017-01-01

Deep learning emerges as a powerful tool for analyzing medical images. Retinal disease detection by using computer-aided diagnosis from fundus image has emerged as a new method. We applied deep learning convolutional neural network by using MatConvNet for an automated detection of multiple retinal diseases with fundus photographs involved in STructured Analysis of the REtina (STARE) database. Dataset was built by expanding data on 10 categories, including normal retina and nine retinal diseases. The optimal outcomes were acquired by using a random forest transfer learning based on VGG-19 architecture. The classification results depended greatly on the number of categories. As the number of categories increased, the performance of deep learning models was diminished. When all 10 categories were included, we obtained results with an accuracy of 30.5%, relative classifier information (RCI) of 0.052, and Cohen’s kappa of 0.224. Considering three integrated normal, background diabetic retinopathy, and dry age-related macular degeneration, the multi-categorical classifier showed accuracy of 72.8%, 0.283 RCI, and 0.577 kappa. In addition, several ensemble classifiers enhanced the multi-categorical classification performance. The transfer learning incorporated with ensemble classifier of clustering and voting approach presented the best performance with accuracy of 36.7%, 0.053 RCI, and 0.225 kappa in the 10 retinal diseases classification problem. First, due to the small size of datasets, the deep learning techniques in this study were ineffective to be applied in clinics where numerous patients suffering from various types of retinal disorders visit for diagnosis and treatment. Second, we found that the transfer learning incorporated with ensemble classifiers can improve the classification performance in order to detect multi-categorical retinal diseases. Further studies should confirm the effectiveness of algorithms with large datasets obtained from hospitals. PMID:29095872

Superpixel-based classification of gastric chromoendoscopy images

NASA Astrophysics Data System (ADS)

Boschetto, Davide; Grisan, Enrico

2017-03-01

Chromoendoscopy (CH) is a gastroenterology imaging modality that involves the staining of tissues with methylene blue, which reacts with the internal walls of the gastrointestinal tract, improving the visual contrast in mucosal surfaces and thus enhancing a doctor's ability to screen precancerous lesions or early cancer. This technique helps identify areas that can be targeted for biopsy or treatment and in this work we will focus on gastric cancer detection. Gastric chromoendoscopy for cancer detection has several taxonomies available, one of which classifies CH images into three classes (normal, metaplasia, dysplasia) based on color, shape and regularity of pit patterns. Computer-assisted diagnosis is desirable to help us improve the reliability of the tissue classification and abnormalities detection. However, traditional computer vision methodologies, mainly segmentation, do not translate well to the specific visual characteristics of a gastroenterology imaging scenario. We propose the exploitation of a first unsupervised segmentation via superpixel, which groups pixels into perceptually meaningful atomic regions, used to replace the rigid structure of the pixel grid. For each superpixel, a set of features is extracted and then fed to a random forest based classifier, which computes a model used to predict the class of each superpixel. The average general accuracy of our model is 92.05% in the pixel domain (86.62% in the superpixel domain), while detection accuracies on the normal and abnormal class are respectively 85.71% and 95%. Eventually, the whole image class can be predicted image through a majority vote on each superpixel's predicted class.

Clustering Single-Cell Expression Data Using Random Forest Graphs.

PubMed

Pouyan, Maziyar Baran; Nourani, Mehrdad

2017-07-01

Complex tissues such as brain and bone marrow are made up of multiple cell types. As the study of biological tissue structure progresses, the role of cell-type-specific research becomes increasingly important. Novel sequencing technology such as single-cell cytometry provides researchers access to valuable biological data. Applying machine-learning techniques to these high-throughput datasets provides deep insights into the cellular landscape of the tissue where those cells are a part of. In this paper, we propose the use of random-forest-based single-cell profiling, a new machine-learning-based technique, to profile different cell types of intricate tissues using single-cell cytometry data. Our technique utilizes random forests to capture cell marker dependences and model the cellular populations using the cell network concept. This cellular network helps us discover what cell types are in the tissue. Our experimental results on public-domain datasets indicate promising performance and accuracy of our technique in extracting cell populations of complex tissues.

Visible and near infrared spectroscopy coupled to random forest to quantify some soil quality parameters

NASA Astrophysics Data System (ADS)

de Santana, Felipe Bachion; de Souza, André Marcelo; Poppi, Ronei Jesus

2018-02-01

This study evaluates the use of visible and near infrared spectroscopy (Vis-NIRS) combined with multivariate regression based on random forest to quantify some quality soil parameters. The parameters analyzed were soil cation exchange capacity (CEC), sum of exchange bases (SB), organic matter (OM), clay and sand present in the soils of several regions of Brazil. Current methods for evaluating these parameters are laborious, timely and require various wet analytical methods that are not adequate for use in precision agriculture, where faster and automatic responses are required. The random forest regression models were statistically better than PLS regression models for CEC, OM, clay and sand, demonstrating resistance to overfitting, attenuating the effect of outlier samples and indicating the most important variables for the model. The methodology demonstrates the potential of the Vis-NIR as an alternative for determination of CEC, SB, OM, sand and clay, making possible to develop a fast and automatic analytical procedure.

Distribution and Characterization of Forested Wetlands in the Carolinas and Virginia

Treesearch

Mark J. Brown

1995-01-01

Recent forest inventories of North Carolina, South Carolina, and Virginia, included sampled for hydric soils, and wetland hydrology. Forest samples that met all 3 of these criteria were classified as forested wetland.This study characterizes wetland forests by extent, owner, age, forest type, physiography, volume, growth, and removals, and evaluates its contribution...

O-GlcNAcPRED-II: an integrated classification algorithm for identifying O-GlcNAcylation sites based on fuzzy undersampling and a K-means PCA oversampling technique.

PubMed

Jia, Cangzhi; Zuo, Yun; Zou, Quan; Hancock, John

2018-02-06

Protein O-GlcNAcylation (O-GlcNAc) is an important post-translational modification of serine (S)/threonine (T) residues that involves multiple molecular and cellular processes. Recent studies have suggested that abnormal O-G1cNAcylation causes many diseases, such as cancer and various neurodegenerative diseases. With the available protein O-G1cNAcylation sites experimentally verified, it is highly desired to develop automated methods to rapidly and effectively identify O-G1cNAcylation sites. Although some computational methods have been proposed, their performance has been unsatisfactory, particularly in terms of prediction sensitivity. In this study, we developed an ensemble model O-GlcNAcPRED-II to identify potential O-G1cNAcylation sites. A K-means principal component analysis oversampling technique (KPCA) and fuzzy undersampling method (FUS) were first proposed and incorporated to reduce the proportion of the original positive and negative training samples. Then, rotation forest, a type of classifier-integrated system, was adopted to divide the eight types of feature space into several subsets using four sub-classifiers: random forest, k-nearest neighbour, naive Bayesian and support vector machine. We observed that O-GlcNAcPRED-II achieved a sensitivity of 81.05%, specificity of 95.91%, accuracy of 91.43% and Matthew's correlation coefficient of 0.7928 for five-fold cross-validation run 10 times. Additionally, the results obtained by O-GlcNAcPRED-II on two independent datasets also indicated that the proposed predictor outperformed five published prediction tools. http://121.42.167.206/OGlcPred/. cangzhijia@dlmu.edu.cn or zouquan@nclab.net. © The Author (2018). Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com

Landsat Based Woody Vegetation Loss Detection in Queensland, Australia Using the Google Earth Engine

NASA Astrophysics Data System (ADS)

Johansen, K.; Phinn, S. R.; Taylor, M.

2014-12-01

Land clearing detection and woody Foliage Projective Cover (FPC) monitoring at the state and national level in Australia has mainly been undertaken by state governments and the Terrestrial Ecosystem Research Network (TERN) because of the considerable expense, expertise, sustained duration of activities and staffing levels needed. Only recently have services become available, providing low budget, generalized access to change detection tools suited to this task. The objective of this research was to examine if a globally available service, Google Earth Engine Beta, could be used to predict woody vegetation loss with accuracies approaching the methods used by TERN and the government of the state of Queensland, Australia. Two change detection approaches were investigated using Landsat Thematic Mapper time series and the Google Earth Engine Application Programming Interface: (1) CART and Random Forest classifiers; and (2) a normalized time series of Foliage Projective Cover (FPC) and NDVI combined with a spectral index. The CART and Random Forest classifiers produced high user's and producer's mapping accuracies of clearing (77-92% and 54-77%, respectively) when detecting change within epochs for which training data were available, but extrapolation to epochs without training data reduced the mapping accuracies. The use of FPC and NDVI time series provided a more robust approach for calculation of a clearing probability, as it did not rely on training data but instead on the difference of the normalized FPC / NDVI mean and standard deviation of a single year at the change point in relation to the remaining time series. However, the FPC and NDVI time series approach represented a trade-off between user's and producer's accuracies. Both change detection approaches explored in this research were sensitive to ephemeral greening and drying of the landscape. However, the developed normalized FPC and NDVI time series approach can be tuned to provide automated alerts for large woody vegetation clearing events by selecting suitable thresholds to identify very likely clearing. This research provides a comprehensive foundation to build further capacity to use globally accessible, free, online image datasets and processing tools to accurately detect woody vegetation clearing in an automated and rapid manner.

Proteus: a random forest classifier to predict disorder-to-order transitioning binding regions in intrinsically disordered proteins

NASA Astrophysics Data System (ADS)

Basu, Sankar; Söderquist, Fredrik; Wallner, Björn

2017-05-01

The focus of the computational structural biology community has taken a dramatic shift over the past one-and-a-half decades from the classical protein structure prediction problem to the possible understanding of intrinsically disordered proteins (IDP) or proteins containing regions of disorder (IDPR). The current interest lies in the unraveling of a disorder-to-order transitioning code embedded in the amino acid sequences of IDPs/IDPRs. Disordered proteins are characterized by an enormous amount of structural plasticity which makes them promiscuous in binding to different partners, multi-functional in cellular activity and atypical in folding energy landscapes resembling partially folded molten globules. Also, their involvement in several deadly human diseases (e.g. cancer, cardiovascular and neurodegenerative diseases) makes them attractive drug targets, and important for a biochemical understanding of the disease(s). The study of the structural ensemble of IDPs is rather difficult, in particular for transient interactions. When bound to a structured partner, an IDPR adapts an ordered conformation in the complex. The residues that undergo this disorder-to-order transition are called protean residues, generally found in short contiguous stretches and the first step in understanding the modus operandi of an IDP/IDPR would be to predict these residues. There are a few available methods which predict these protean segments from their amino acid sequences; however, their performance reported in the literature leaves clear room for improvement. With this background, the current study presents `Proteus', a random forest classifier that predicts the likelihood of a residue undergoing a disorder-to-order transition upon binding to a potential partner protein. The prediction is based on features that can be calculated using the amino acid sequence alone. Proteus compares favorably with existing methods predicting twice as many true positives as the second best method (55 vs. 27%) with a much higher precision on an independent data set. The current study also sheds some light on a possible `disorder-to-order' transitioning consensus, untangled, yet embedded in the amino acid sequence of IDPs. Some guidelines have also been suggested for proceeding with a real-life structural modeling involving an IDPR using Proteus.

Random forest wetland classification using ALOS-2 L-band, RADARSAT-2 C-band, and TerraSAR-X imagery

NASA Astrophysics Data System (ADS)

Mahdianpari, Masoud; Salehi, Bahram; Mohammadimanesh, Fariba; Motagh, Mahdi

2017-08-01

Wetlands are important ecosystems around the world, although they are degraded due both to anthropogenic and natural process. Newfoundland is among the richest Canadian province in terms of different wetland classes. Herbaceous wetlands cover extensive areas of the Avalon Peninsula, which are the habitat of a number of animal and plant species. In this study, a novel hierarchical object-based Random Forest (RF) classification approach is proposed for discriminating between different wetland classes in a sub-region located in the north eastern portion of the Avalon Peninsula. Particularly, multi-polarization and multi-frequency SAR data, including X-band TerraSAR-X single polarized (HH), L-band ALOS-2 dual polarized (HH/HV), and C-band RADARSAT-2 fully polarized images, were applied in different classification levels. First, a SAR backscatter analysis of different land cover types was performed by training data and used in Level-I classification to separate water from non-water classes. This was followed by Level-II classification, wherein the water class was further divided into shallow- and deep-water classes, and the non-water class was partitioned into herbaceous and non-herbaceous classes. In Level-III classification, the herbaceous class was further divided into bog, fen, and marsh classes, while the non-herbaceous class was subsequently partitioned into urban, upland, and swamp classes. In Level-II and -III classifications, different polarimetric decomposition approaches, including Cloude-Pottier, Freeman-Durden, Yamaguchi decompositions, and Kennaugh matrix elements were extracted to aid the RF classifier. The overall accuracy and kappa coefficient were determined in each classification level for evaluating the classification results. The importance of input features was also determined using the variable importance obtained by RF. It was found that the Kennaugh matrix elements, Yamaguchi, and Freeman-Durden decompositions were the most important parameters for wetland classification in this study. Using this new hierarchical RF classification approach, an overall accuracy of up to 94% was obtained for classifying different land cover types in the study area.

Impact Assessment of Mikania Micrantha on Land Cover and Maxent Modeling to Predict its Potential Invasion Sites

NASA Astrophysics Data System (ADS)

Baidar, T.; Shrestha, A. B.; Ranjit, R.; Adhikari, R.; Ghimire, S.; Shrestha, N.

2017-05-01

Mikania micrantha is one of the major invasive alien plant species in tropical moist forest regions of Asia including Nepal. Recently, this weed is spreading at an alarming rate in Chitwan National Park (CNP) and threatening biodiversity. This paper aims to assess the impacts of Mikania micrantha on different land cover and to predict potential invasion sites in CNP using Maxent model. Primary data for this were presence point coordinates and perceived Mikania micrantha cover collected through systematic random sampling technique. Rapideye image, Shuttle Radar Topographic Mission data and bioclimatic variables were acquired as secondary data. Mikania micrantha distribution maps were prepared by overlaying the presence points on image classified by object based image analysis. The overall accuracy of classification was 90 % with Kappa coefficient 0.848. A table depicting the number of sample points in each land cover with respective Mikania micrantha coverage was extracted from the distribution maps to show the impact. The riverine forest was found to be the most affected land cover with 85.98 % presence points and sal forest was found to be very less affected with only 17.02 % presence points. Maxent modeling predicted the areas near the river valley as the potential invasion sites with statistically significant Area Under the Receiver Operating Curve (AUC) value of 0.969. Maximum temperature of warmest month and annual precipitation were identified as the predictor variables that contribute the most to Mikania micrantha's potential distribution.

rpiCOOL: A tool for In Silico RNA-protein interaction detection using random forest.

PubMed

Akbaripour-Elahabad, Mohammad; Zahiri, Javad; Rafeh, Reza; Eslami, Morteza; Azari, Mahboobeh

2016-08-07

Understanding the principle of RNA-protein interactions (RPIs) is of critical importance to provide insights into post-transcriptional gene regulation and is useful to guide studies about many complex diseases. The limitations and difficulties associated with experimental determination of RPIs, call an urgent need to computational methods for RPI prediction. In this paper, we proposed a machine learning method to detect RNA-protein interactions based on sequence information. We used motif information and repetitive patterns, which have been extracted from experimentally validated RNA-protein interactions, in combination with sequence composition as descriptors to build a model to RPI prediction via a random forest classifier. About 20% of the "sequence motifs" and "nucleotide composition" features have been selected as the informative features with the feature selection methods. These results suggest that these two feature types contribute effectively in RPI detection. Results of 10-fold cross-validation experiments on three non-redundant benchmark datasets show a better performance of the proposed method in comparison with the current state-of-the-art methods in terms of various performance measures. In addition, the results revealed that the accuracy of the RPI prediction methods could vary considerably across different organisms. We have implemented the proposed method, namely rpiCOOL, as a stand-alone tool with a user friendly graphical user interface (GUI) that enables the researchers to predict RNA-protein interaction. The rpiCOOL is freely available at http://biocool.ir/rpicool.html for non-commercial uses. Copyright © 2016 Elsevier Ltd. All rights reserved.

The ASAS-SN Catalog of Variable Stars I: The Serendipitous Survey

NASA Astrophysics Data System (ADS)

Jayasinghe, T.; Kochanek, C. S.; Stanek, K. Z.; Shappee, B. J.; Holoien, T. W.-S.; Thompson, Todd A.; Prieto, J. L.; Dong, Subo; Pawlak, M.; Shields, J. V.; Pojmanski, G.; Otero, S.; Britt, C. A.; Will, D.

2018-04-01

The All-Sky Automated Survey for Supernovae (ASAS-SN) is the first optical survey to routinely monitor the whole sky with a cadence of ˜2 - 3 days down to V≲ 17 mag. ASAS-SN has monitored the whole sky since 2014, collecting ˜100 - 500 epochs of observations per field. The V-band light curves for candidate variables identified during the search for supernovae are classified using a random forest classifier and visually verified. We present a catalog of 66,533 bright, new variable stars discovered during our search for supernovae, including 27,753 periodic variables and 38,780 irregular variables. V-band light curves for the ASAS-SN variables are available through the ASAS-SN variable stars database (https://asas-sn.osu.edu/variables). The database will begin to include the light curves of known variable stars in the near future along with the results for a systematic, all-sky variability survey.

Deep learning approach to bacterial colony classification.

PubMed

Zieliński, Bartosz; Plichta, Anna; Misztal, Krzysztof; Spurek, Przemysław; Brzychczy-Włoch, Monika; Ochońska, Dorota

2017-01-01

In microbiology it is diagnostically useful to recognize various genera and species of bacteria. It can be achieved using computer-aided methods, which make the recognition processes more automatic and thus significantly reduce the time necessary for the classification. Moreover, in case of diagnostic uncertainty (the misleading similarity in shape or structure of bacterial cells), such methods can minimize the risk of incorrect recognition. In this article, we apply the state of the art method for texture analysis to classify genera and species of bacteria. This method uses deep Convolutional Neural Networks to obtain image descriptors, which are then encoded and classified with Support Vector Machine or Random Forest. To evaluate this approach and to make it comparable with other approaches, we provide a new dataset of images. DIBaS dataset (Digital Image of Bacterial Species) contains 660 images with 33 different genera and species of bacteria.

Interpreting Black-Box Classifiers Using Instance-Level Visual Explanations

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

Tamagnini, Paolo; Krause, Josua W.; Dasgupta, Aritra

2017-05-14

To realize the full potential of machine learning in diverse real- world domains, it is necessary for model predictions to be readily interpretable and actionable for the human in the loop. Analysts, who are the users but not the developers of machine learning models, often do not trust a model because of the lack of transparency in associating predictions with the underlying data space. To address this problem, we propose Rivelo, a visual analytic interface that enables analysts to understand the causes behind predictions of binary classifiers by interactively exploring a set of instance-level explanations. These explanations are model-agnostic, treatingmore » a model as a black box, and they help analysts in interactively probing the high-dimensional binary data space for detecting features relevant to predictions. We demonstrate the utility of the interface with a case study analyzing a random forest model on the sentiment of Yelp reviews about doctors.« less

Pressure sensor based on pristine multi-walled carbon nanotubes forest

NASA Astrophysics Data System (ADS)

Yasar, M.; Mohamed, N. M.; Hamid, N. H.; Shuaib, M.

2016-11-01

In the course of the most recent decade, carbon nanotubes (CNTs) have been developed as alternate material for many sensing applications because of their interesting properties. Their outstanding electromechanical properties make them suitable for pressure/strain sensing application. Other than in view of their structure and number of walls (i.e. Single-Walled CNTs and MultiWalled CNTs), carbon nanotubes can likewise be classified based on their orientation and combined arrangement. One such classification is vertically aligned Multi-Walled Carbon Nanotubes (VA-MWCNTs), regularly termed as CNTs arrays, foam or forest which is macro scale form of CNTs. Elastic behavior alongside exceptional electromechanical (high gauge factor) make it suitable for pressure sensing applications. This paper presents pressure sensor based on such carbon nanotubes forest in pristine form which enables it to perform over wider temperature range as compared to pressure sensors based on conventional materials such as Silicon.

Lidar-based multinomial classification algorithms for tropical forest degradation status: Implications for biomass estimation

NASA Astrophysics Data System (ADS)

Duffy, P.; Keller, M.; Longo, M.; Morton, D. C.; dos-Santos, M. N.; Pinagé, E. R.

2017-12-01

There is an urgent need to quantify the effects of land use and land cover change on carbon stocks in tropical forests to support REDD+ policies and improve characterization of global carbon budgets. This need is underscored by the fact that the variability in forest biomass estimates from global forest carbon maps is artificially low relative to estimates generated from forest inventory and high-resolution airborne lidar data. Both deforestation and degradation processes (e.g. logging, fire, and fragmentation) affect carbon fluxes at varying spatial and temporal scales. While the spatial extent and impact of deforestation has been relatively well characterized, the quantification of degradation processes is still poorly constrained. In the Brazilian Amazon, the largest source of uncertainty in CO2 emissions estimates is data on changes in tropical forest carbon stocks through time, followed closely by incomplete information on the carbon losses from forest degradation. In this work, we present a method for classifying the degradation status of tropical forests using higher order moments (skewness and kurtosis) of lidar return distributions aggregated at grids with resolution ranging from 50 m to 250 m. Across multiple spatial resolutions, we quantify the strength of the functional relationship between the lidar returns and the classification based on historical time series of Landsat imagery. Our results show that the higher order moments of the lidar return distributions provide sufficient information to build multinomial models that accurately classify the landscape into intact, logged, and burned forests. Model fit improved with coarser spatial resolution with Kappa statistics of 0.70 at 50 m, and 0.77 at 250 m. In addition, multi-class AUC was estimated as 0.87 at 50 m, and 0.95 at 250 m. This classification provides important information regarding the applicability of the use of lidar data for regional monitoring of recent logging, as well as the trajectory of the carbon budget. Differentiating between the biomass changes associated with deforestation and degradation processes is critical for accurate accounting of disturbance impacts on carbon cycling within the Brazilian Amazon and global tropical forests.

Estimation of sleep status in sleep apnea patients using a novel head actigraphy technique.

PubMed

Hummel, Richard; Bradley, T Douglas; Fernie, Geoff R; Chang, S J Isaac; Alshaer, Hisham

2015-01-01

Polysomnography is a comprehensive modality for diagnosing sleep apnea (SA), but it is expensive and not widely available. Several technologies have been developed for portable diagnosis of SA in the home, most of which lack the ability to detect sleep status. Wrist actigraphy (accelerometry) has been adopted to cover this limitation. However, head actigraphy has not been systematically evaluated for this purpose. Therefore, the aim of this study was to evaluate the ability of head actigraphy to detect sleep/wake status. We obtained full overnight 3-axis head accelerometry data from 75 sleep apnea patient recordings. These were split into training and validation groups (2:1). Data were preprocessed and 5 features were extracted. Different feature combinations were fed into 3 different classifiers, namely support vector machine, logistic regression, and random forests, each of which was trained and validated on a different subgroup. The random forest algorithm yielded the highest performance, with an area under the receiver operating characteristic (ROC) curve of 0.81 for detection of sleep status. This shows that this technique has a very good performance in detecting sleep status in SA patients despite the specificities in this population, such as respiration related movements.