Sparse alignment for robust tensor learning.

PubMed

Lai, Zhihui; Wong, Wai Keung; Xu, Yong; Zhao, Cairong; Sun, Mingming

2014-10-01

Multilinear/tensor extensions of manifold learning based algorithms have been widely used in computer vision and pattern recognition. This paper first provides a systematic analysis of the multilinear extensions for the most popular methods by using alignment techniques, thereby obtaining a general tensor alignment framework. From this framework, it is easy to show that the manifold learning based tensor learning methods are intrinsically different from the alignment techniques. Based on the alignment framework, a robust tensor learning method called sparse tensor alignment (STA) is then proposed for unsupervised tensor feature extraction. Different from the existing tensor learning methods, L1- and L2-norms are introduced to enhance the robustness in the alignment step of the STA. The advantage of the proposed technique is that the difficulty in selecting the size of the local neighborhood can be avoided in the manifold learning based tensor feature extraction algorithms. Although STA is an unsupervised learning method, the sparsity encodes the discriminative information in the alignment step and provides the robustness of STA. Extensive experiments on the well-known image databases as well as action and hand gesture databases by encoding object images as tensors demonstrate that the proposed STA algorithm gives the most competitive performance when compared with the tensor-based unsupervised learning methods.

Unsupervised Feature Learning for Heart Sounds Classification Using Autoencoder

NASA Astrophysics Data System (ADS)

Hu, Wei; Lv, Jiancheng; Liu, Dongbo; Chen, Yao

2018-04-01

Cardiovascular disease seriously threatens the health of many people. It is usually diagnosed during cardiac auscultation, which is a fast and efficient method of cardiovascular disease diagnosis. In recent years, deep learning approach using unsupervised learning has made significant breakthroughs in many fields. However, to our knowledge, deep learning has not yet been used for heart sound classification. In this paper, we first use the average Shannon energy to extract the envelope of the heart sounds, then find the highest point of S1 to extract the cardiac cycle. We convert the time-domain signals of the cardiac cycle into spectrograms and apply principal component analysis whitening to reduce the dimensionality of the spectrogram. Finally, we apply a two-layer autoencoder to extract the features of the spectrogram. The experimental results demonstrate that the features from the autoencoder are suitable for heart sound classification.

MARTA GANs: Unsupervised Representation Learning for Remote Sensing Image Classification

NASA Astrophysics Data System (ADS)

Lin, Daoyu; Fu, Kun; Wang, Yang; Xu, Guangluan; Sun, Xian

2017-11-01

With the development of deep learning, supervised learning has frequently been adopted to classify remotely sensed images using convolutional networks (CNNs). However, due to the limited amount of labeled data available, supervised learning is often difficult to carry out. Therefore, we proposed an unsupervised model called multiple-layer feature-matching generative adversarial networks (MARTA GANs) to learn a representation using only unlabeled data. MARTA GANs consists of both a generative model $G$ and a discriminative model $D$. We treat $D$ as a feature extractor. To fit the complex properties of remote sensing data, we use a fusion layer to merge the mid-level and global features. $G$ can produce numerous images that are similar to the training data; therefore, $D$ can learn better representations of remotely sensed images using the training data provided by $G$. The classification results on two widely used remote sensing image databases show that the proposed method significantly improves the classification performance compared with other state-of-the-art methods.

Multi-Source Multi-Target Dictionary Learning for Prediction of Cognitive Decline

PubMed Central

Zhang, Jie; Li, Qingyang; Caselli, Richard J.; Thompson, Paul M.; Ye, Jieping; Wang, Yalin

2017-01-01

Alzheimer’s Disease (AD) is the most common type of dementia. Identifying correct biomarkers may determine pre-symptomatic AD subjects and enable early intervention. Recently, Multi-task sparse feature learning has been successfully applied to many computer vision and biomedical informatics researches. It aims to improve the generalization performance by exploiting the shared features among different tasks. However, most of the existing algorithms are formulated as a supervised learning scheme. Its drawback is with either insufficient feature numbers or missing label information. To address these challenges, we formulate an unsupervised framework for multi-task sparse feature learning based on a novel dictionary learning algorithm. To solve the unsupervised learning problem, we propose a two-stage Multi-Source Multi-Target Dictionary Learning (MMDL) algorithm. In stage 1, we propose a multi-source dictionary learning method to utilize the common and individual sparse features in different time slots. In stage 2, supported by a rigorous theoretical analysis, we develop a multi-task learning method to solve the missing label problem. Empirical studies on an N = 3970 longitudinal brain image data set, which involves 2 sources and 5 targets, demonstrate the improved prediction accuracy and speed efficiency of MMDL in comparison with other state-of-the-art algorithms. PMID:28943731

Prediction task guided representation learning of medical codes in EHR.

PubMed

Cui, Liwen; Xie, Xiaolei; Shen, Zuojun

2018-06-18

There have been rapidly growing applications using machine learning models for predictive analytics in Electronic Health Records (EHR) to improve the quality of hospital services and the efficiency of healthcare resource utilization. A fundamental and crucial step in developing such models is to convert medical codes in EHR to feature vectors. These medical codes are used to represent diagnoses or procedures. Their vector representations have a tremendous impact on the performance of machine learning models. Recently, some researchers have utilized representation learning methods from Natural Language Processing (NLP) to learn vector representations of medical codes. However, most previous approaches are unsupervised, i.e. the generation of medical code vectors is independent from prediction tasks. Thus, the obtained feature vectors may be inappropriate for a specific prediction task. Moreover, unsupervised methods often require a lot of samples to obtain reliable results, but most practical problems have very limited patient samples. In this paper, we develop a new method called Prediction Task Guided Health Record Aggregation (PTGHRA), which aggregates health records guided by prediction tasks, to construct training corpus for various representation learning models. Compared with unsupervised approaches, representation learning models integrated with PTGHRA yield a significant improvement in predictive capability of generated medical code vectors, especially for limited training samples. Copyright © 2018. Published by Elsevier Inc.

Age and gender classification in the wild with unsupervised feature learning

NASA Astrophysics Data System (ADS)

Wan, Lihong; Huo, Hong; Fang, Tao

2017-03-01

Inspired by unsupervised feature learning (UFL) within the self-taught learning framework, we propose a method based on UFL, convolution representation, and part-based dimensionality reduction to handle facial age and gender classification, which are two challenging problems under unconstrained circumstances. First, UFL is introduced to learn selective receptive fields (filters) automatically by applying whitening transformation and spherical k-means on random patches collected from unlabeled data. The learning process is fast and has no hyperparameters to tune. Then, the input image is convolved with these filters to obtain filtering responses on which local contrast normalization is applied. Average pooling and feature concatenation are then used to form global face representation. Finally, linear discriminant analysis with part-based strategy is presented to reduce the dimensions of the global representation and to improve classification performances further. Experiments on three challenging databases, namely, Labeled faces in the wild, Gallagher group photos, and Adience, demonstrate the effectiveness of the proposed method relative to that of state-of-the-art approaches.

Unsupervised spike sorting based on discriminative subspace learning.

PubMed

Keshtkaran, Mohammad Reza; Yang, Zhi

2014-01-01

Spike sorting is a fundamental preprocessing step for many neuroscience studies which rely on the analysis of spike trains. In this paper, we present two unsupervised spike sorting algorithms based on discriminative subspace learning. The first algorithm simultaneously learns the discriminative feature subspace and performs clustering. It uses histogram of features in the most discriminative projection to detect the number of neurons. The second algorithm performs hierarchical divisive clustering that learns a discriminative 1-dimensional subspace for clustering in each level of the hierarchy until achieving almost unimodal distribution in the subspace. The algorithms are tested on synthetic and in-vivo data, and are compared against two widely used spike sorting methods. The comparative results demonstrate that our spike sorting methods can achieve substantially higher accuracy in lower dimensional feature space, and they are highly robust to noise. Moreover, they provide significantly better cluster separability in the learned subspace than in the subspace obtained by principal component analysis or wavelet transform.

Enhanced HMAX model with feedforward feature learning for multiclass categorization.

PubMed

Li, Yinlin; Wu, Wei; Zhang, Bo; Li, Fengfu

2015-01-01

In recent years, the interdisciplinary research between neuroscience and computer vision has promoted the development in both fields. Many biologically inspired visual models are proposed, and among them, the Hierarchical Max-pooling model (HMAX) is a feedforward model mimicking the structures and functions of V1 to posterior inferotemporal (PIT) layer of the primate visual cortex, which could generate a series of position- and scale- invariant features. However, it could be improved with attention modulation and memory processing, which are two important properties of the primate visual cortex. Thus, in this paper, based on recent biological research on the primate visual cortex, we still mimic the first 100-150 ms of visual cognition to enhance the HMAX model, which mainly focuses on the unsupervised feedforward feature learning process. The main modifications are as follows: (1) To mimic the attention modulation mechanism of V1 layer, a bottom-up saliency map is computed in the S1 layer of the HMAX model, which can support the initial feature extraction for memory processing; (2) To mimic the learning, clustering and short-term memory to long-term memory conversion abilities of V2 and IT, an unsupervised iterative clustering method is used to learn clusters with multiscale middle level patches, which are taken as long-term memory; (3) Inspired by the multiple feature encoding mode of the primate visual cortex, information including color, orientation, and spatial position are encoded in different layers of the HMAX model progressively. By adding a softmax layer at the top of the model, multiclass categorization experiments can be conducted, and the results on Caltech101 show that the enhanced model with a smaller memory size exhibits higher accuracy than the original HMAX model, and could also achieve better accuracy than other unsupervised feature learning methods in multiclass categorization task.

Computational phenotype discovery using unsupervised feature learning over noisy, sparse, and irregular clinical data.

PubMed

Lasko, Thomas A; Denny, Joshua C; Levy, Mia A

2013-01-01

Inferring precise phenotypic patterns from population-scale clinical data is a core computational task in the development of precision, personalized medicine. The traditional approach uses supervised learning, in which an expert designates which patterns to look for (by specifying the learning task and the class labels), and where to look for them (by specifying the input variables). While appropriate for individual tasks, this approach scales poorly and misses the patterns that we don't think to look for. Unsupervised feature learning overcomes these limitations by identifying patterns (or features) that collectively form a compact and expressive representation of the source data, with no need for expert input or labeled examples. Its rising popularity is driven by new deep learning methods, which have produced high-profile successes on difficult standardized problems of object recognition in images. Here we introduce its use for phenotype discovery in clinical data. This use is challenging because the largest source of clinical data - Electronic Medical Records - typically contains noisy, sparse, and irregularly timed observations, rendering them poor substrates for deep learning methods. Our approach couples dirty clinical data to deep learning architecture via longitudinal probability densities inferred using Gaussian process regression. From episodic, longitudinal sequences of serum uric acid measurements in 4368 individuals we produced continuous phenotypic features that suggest multiple population subtypes, and that accurately distinguished (0.97 AUC) the uric-acid signatures of gout vs. acute leukemia despite not being optimized for the task. The unsupervised features were as accurate as gold-standard features engineered by an expert with complete knowledge of the domain, the classification task, and the class labels. Our findings demonstrate the potential for achieving computational phenotype discovery at population scale. We expect such data-driven phenotypes to expose unknown disease variants and subtypes and to provide rich targets for genetic association studies.

Computational Phenotype Discovery Using Unsupervised Feature Learning over Noisy, Sparse, and Irregular Clinical Data

PubMed Central

Lasko, Thomas A.; Denny, Joshua C.; Levy, Mia A.

2013-01-01

Inferring precise phenotypic patterns from population-scale clinical data is a core computational task in the development of precision, personalized medicine. The traditional approach uses supervised learning, in which an expert designates which patterns to look for (by specifying the learning task and the class labels), and where to look for them (by specifying the input variables). While appropriate for individual tasks, this approach scales poorly and misses the patterns that we don’t think to look for. Unsupervised feature learning overcomes these limitations by identifying patterns (or features) that collectively form a compact and expressive representation of the source data, with no need for expert input or labeled examples. Its rising popularity is driven by new deep learning methods, which have produced high-profile successes on difficult standardized problems of object recognition in images. Here we introduce its use for phenotype discovery in clinical data. This use is challenging because the largest source of clinical data – Electronic Medical Records – typically contains noisy, sparse, and irregularly timed observations, rendering them poor substrates for deep learning methods. Our approach couples dirty clinical data to deep learning architecture via longitudinal probability densities inferred using Gaussian process regression. From episodic, longitudinal sequences of serum uric acid measurements in 4368 individuals we produced continuous phenotypic features that suggest multiple population subtypes, and that accurately distinguished (0.97 AUC) the uric-acid signatures of gout vs. acute leukemia despite not being optimized for the task. The unsupervised features were as accurate as gold-standard features engineered by an expert with complete knowledge of the domain, the classification task, and the class labels. Our findings demonstrate the potential for achieving computational phenotype discovery at population scale. We expect such data-driven phenotypes to expose unknown disease variants and subtypes and to provide rich targets for genetic association studies. PMID:23826094

Analyzing Distributional Learning of Phonemic Categories in Unsupervised Deep Neural Networks

PubMed Central

Räsänen, Okko; Nagamine, Tasha; Mesgarani, Nima

2017-01-01

Infants’ speech perception adapts to the phonemic categories of their native language, a process assumed to be driven by the distributional properties of speech. This study investigates whether deep neural networks (DNNs), the current state-of-the-art in distributional feature learning, are capable of learning phoneme-like representations of speech in an unsupervised manner. We trained DNNs with unlabeled and labeled speech and analyzed the activations of each layer with respect to the phones in the input segments. The analyses reveal that the emergence of phonemic invariance in DNNs is dependent on the availability of phonemic labeling of the input during the training. No increased phonemic selectivity of the hidden layers was observed in the purely unsupervised networks despite successful learning of low-dimensional representations for speech. This suggests that additional learning constraints or more sophisticated models are needed to account for the emergence of phone-like categories in distributional learning operating on natural speech. PMID:29359204

Comparative study of feature selection with ensemble learning using SOM variants

NASA Astrophysics Data System (ADS)

Filali, Ameni; Jlassi, Chiraz; Arous, Najet

2017-03-01

Ensemble learning has succeeded in the growth of stability and clustering accuracy, but their runtime prohibits them from scaling up to real-world applications. This study deals the problem of selecting a subset of the most pertinent features for every cluster from a dataset. The proposed method is another extension of the Random Forests approach using self-organizing maps (SOM) variants to unlabeled data that estimates the out-of-bag feature importance from a set of partitions. Every partition is created using a various bootstrap sample and a random subset of the features. Then, we show that the process internal estimates are used to measure variable pertinence in Random Forests are also applicable to feature selection in unsupervised learning. This approach aims to the dimensionality reduction, visualization and cluster characterization at the same time. Hence, we provide empirical results on nineteen benchmark data sets indicating that RFS can lead to significant improvement in terms of clustering accuracy, over several state-of-the-art unsupervised methods, with a very limited subset of features. The approach proves promise to treat with very broad domains.

Segmentation of magnetic resonance images using fuzzy algorithms for learning vector quantization.

PubMed

Karayiannis, N B; Pai, P I

1999-02-01

This paper evaluates a segmentation technique for magnetic resonance (MR) images of the brain based on fuzzy algorithms for learning vector quantization (FALVQ). These algorithms perform vector quantization by updating all prototypes of a competitive network through an unsupervised learning process. Segmentation of MR images is formulated as an unsupervised vector quantization process, where the local values of different relaxation parameters form the feature vectors which are represented by a relatively small set of prototypes. The experiments evaluate a variety of FALVQ algorithms in terms of their ability to identify different tissues and discriminate between normal tissues and abnormalities.

Probability density function learning by unsupervised neurons.

PubMed

Fiori, S

2001-10-01

In a recent work, we introduced the concept of pseudo-polynomial adaptive activation function neuron (FAN) and presented an unsupervised information-theoretic learning theory for such structure. The learning model is based on entropy optimization and provides a way of learning probability distributions from incomplete data. The aim of the present paper is to illustrate some theoretical features of the FAN neuron, to extend its learning theory to asymmetrical density function approximation, and to provide an analytical and numerical comparison with other known density function estimation methods, with special emphasis to the universal approximation ability. The paper also provides a survey of PDF learning from incomplete data, as well as results of several experiments performed on real-world problems and signals.

Extraction of temporally correlated features from dynamic vision sensors with spike-timing-dependent plasticity.

PubMed

Bichler, Olivier; Querlioz, Damien; Thorpe, Simon J; Bourgoin, Jean-Philippe; Gamrat, Christian

2012-08-01

A biologically inspired approach to learning temporally correlated patterns from a spiking silicon retina is presented. Spikes are generated from the retina in response to relative changes in illumination at the pixel level and transmitted to a feed-forward spiking neural network. Neurons become sensitive to patterns of pixels with correlated activation times, in a fully unsupervised scheme. This is achieved using a special form of Spike-Timing-Dependent Plasticity which depresses synapses that did not recently contribute to the post-synaptic spike activation, regardless of their activation time. Competitive learning is implemented with lateral inhibition. When tested with real-life data, the system is able to extract complex and overlapping temporally correlated features such as car trajectories on a freeway, after only 10 min of traffic learning. Complete trajectories can be learned with a 98% detection rate using a second layer, still with unsupervised learning, and the system may be used as a car counter. The proposed neural network is extremely robust to noise and it can tolerate a high degree of synaptic and neuronal variability with little impact on performance. Such results show that a simple biologically inspired unsupervised learning scheme is capable of generating selectivity to complex meaningful events on the basis of relatively little sensory experience. Copyright © 2012 Elsevier Ltd. All rights reserved.

Video mining using combinations of unsupervised and supervised learning techniques

NASA Astrophysics Data System (ADS)

Divakaran, Ajay; Miyahara, Koji; Peker, Kadir A.; Radhakrishnan, Regunathan; Xiong, Ziyou

2003-12-01

We discuss the meaning and significance of the video mining problem, and present our work on some aspects of video mining. A simple definition of video mining is unsupervised discovery of patterns in audio-visual content. Such purely unsupervised discovery is readily applicable to video surveillance as well as to consumer video browsing applications. We interpret video mining as content-adaptive or "blind" content processing, in which the first stage is content characterization and the second stage is event discovery based on the characterization obtained in stage 1. We discuss the target applications and find that using a purely unsupervised approach are too computationally complex to be implemented on our product platform. We then describe various combinations of unsupervised and supervised learning techniques that help discover patterns that are useful to the end-user of the application. We target consumer video browsing applications such as commercial message detection, sports highlights extraction etc. We employ both audio and video features. We find that supervised audio classification combined with unsupervised unusual event discovery enables accurate supervised detection of desired events. Our techniques are computationally simple and robust to common variations in production styles etc.

Communications and control for electric power systems: Power flow classification for static security assessment

NASA Technical Reports Server (NTRS)

Niebur, D.; Germond, A.

1993-01-01

This report investigates the classification of power system states using an artificial neural network model, Kohonen's self-organizing feature map. The ultimate goal of this classification is to assess power system static security in real-time. Kohonen's self-organizing feature map is an unsupervised neural network which maps N-dimensional input vectors to an array of M neurons. After learning, the synaptic weight vectors exhibit a topological organization which represents the relationship between the vectors of the training set. This learning is unsupervised, which means that the number and size of the classes are not specified beforehand. In the application developed in this report, the input vectors used as the training set are generated by off-line load-flow simulations. The learning algorithm and the results of the organization are discussed.

Space Object Classification Using Fused Features of Time Series Data

NASA Astrophysics Data System (ADS)

Jia, B.; Pham, K. D.; Blasch, E.; Shen, D.; Wang, Z.; Chen, G.

In this paper, a fused feature vector consisting of raw time series and texture feature information is proposed for space object classification. The time series data includes historical orbit trajectories and asteroid light curves. The texture feature is derived from recurrence plots using Gabor filters for both unsupervised learning and supervised learning algorithms. The simulation results show that the classification algorithms using the fused feature vector achieve better performance than those using raw time series or texture features only.

Weakly supervised visual dictionary learning by harnessing image attributes.

PubMed

Gao, Yue; Ji, Rongrong; Liu, Wei; Dai, Qionghai; Hua, Gang

2014-12-01

Bag-of-features (BoFs) representation has been extensively applied to deal with various computer vision applications. To extract discriminative and descriptive BoF, one important step is to learn a good dictionary to minimize the quantization loss between local features and codewords. While most existing visual dictionary learning approaches are engaged with unsupervised feature quantization, the latest trend has turned to supervised learning by harnessing the semantic labels of images or regions. However, such labels are typically too expensive to acquire, which restricts the scalability of supervised dictionary learning approaches. In this paper, we propose to leverage image attributes to weakly supervise the dictionary learning procedure without requiring any actual labels. As a key contribution, our approach establishes a generative hidden Markov random field (HMRF), which models the quantized codewords as the observed states and the image attributes as the hidden states, respectively. Dictionary learning is then performed by supervised grouping the observed states, where the supervised information is stemmed from the hidden states of the HMRF. In such a way, the proposed dictionary learning approach incorporates the image attributes to learn a semantic-preserving BoF representation without any genuine supervision. Experiments in large-scale image retrieval and classification tasks corroborate that our approach significantly outperforms the state-of-the-art unsupervised dictionary learning approaches.

Discrete Wavelet Transform-Based Whole-Spectral and Subspectral Analysis for Improved Brain Tumor Clustering Using Single Voxel MR Spectroscopy.

PubMed

Yang, Guang; Nawaz, Tahir; Barrick, Thomas R; Howe, Franklyn A; Slabaugh, Greg

2015-12-01

Many approaches have been considered for automatic grading of brain tumors by means of pattern recognition with magnetic resonance spectroscopy (MRS). Providing an improved technique which can assist clinicians in accurately identifying brain tumor grades is our main objective. The proposed technique, which is based on the discrete wavelet transform (DWT) of whole-spectral or subspectral information of key metabolites, combined with unsupervised learning, inspects the separability of the extracted wavelet features from the MRS signal to aid the clustering. In total, we included 134 short echo time single voxel MRS spectra (SV MRS) in our study that cover normal controls, low grade and high grade tumors. The combination of DWT-based whole-spectral or subspectral analysis and unsupervised clustering achieved an overall clustering accuracy of 94.8% and a balanced error rate of 7.8%. To the best of our knowledge, it is the first study using DWT combined with unsupervised learning to cluster brain SV MRS. Instead of dimensionality reduction on SV MRS or feature selection using model fitting, our study provides an alternative method of extracting features to obtain promising clustering results.

Scalable High Performance Image Registration Framework by Unsupervised Deep Feature Representations Learning

PubMed Central

Wu, Guorong; Kim, Minjeong; Wang, Qian; Munsell, Brent C.

2015-01-01

Feature selection is a critical step in deformable image registration. In particular, selecting the most discriminative features that accurately and concisely describe complex morphological patterns in image patches improves correspondence detection, which in turn improves image registration accuracy. Furthermore, since more and more imaging modalities are being invented to better identify morphological changes in medical imaging data,, the development of deformable image registration method that scales well to new image modalities or new image applications with little to no human intervention would have a significant impact on the medical image analysis community. To address these concerns, a learning-based image registration framework is proposed that uses deep learning to discover compact and highly discriminative features upon observed imaging data. Specifically, the proposed feature selection method uses a convolutional stacked auto-encoder to identify intrinsic deep feature representations in image patches. Since deep learning is an unsupervised learning method, no ground truth label knowledge is required. This makes the proposed feature selection method more flexible to new imaging modalities since feature representations can be directly learned from the observed imaging data in a very short amount of time. Using the LONI and ADNI imaging datasets, image registration performance was compared to two existing state-of-the-art deformable image registration methods that use handcrafted features. To demonstrate the scalability of the proposed image registration framework image registration experiments were conducted on 7.0-tesla brain MR images. In all experiments, the results showed the new image registration framework consistently demonstrated more accurate registration results when compared to state-of-the-art. PMID:26552069

Scalable High-Performance Image Registration Framework by Unsupervised Deep Feature Representations Learning.

PubMed

Wu, Guorong; Kim, Minjeong; Wang, Qian; Munsell, Brent C; Shen, Dinggang

2016-07-01

Feature selection is a critical step in deformable image registration. In particular, selecting the most discriminative features that accurately and concisely describe complex morphological patterns in image patches improves correspondence detection, which in turn improves image registration accuracy. Furthermore, since more and more imaging modalities are being invented to better identify morphological changes in medical imaging data, the development of deformable image registration method that scales well to new image modalities or new image applications with little to no human intervention would have a significant impact on the medical image analysis community. To address these concerns, a learning-based image registration framework is proposed that uses deep learning to discover compact and highly discriminative features upon observed imaging data. Specifically, the proposed feature selection method uses a convolutional stacked autoencoder to identify intrinsic deep feature representations in image patches. Since deep learning is an unsupervised learning method, no ground truth label knowledge is required. This makes the proposed feature selection method more flexible to new imaging modalities since feature representations can be directly learned from the observed imaging data in a very short amount of time. Using the LONI and ADNI imaging datasets, image registration performance was compared to two existing state-of-the-art deformable image registration methods that use handcrafted features. To demonstrate the scalability of the proposed image registration framework, image registration experiments were conducted on 7.0-T brain MR images. In all experiments, the results showed that the new image registration framework consistently demonstrated more accurate registration results when compared to state of the art.

Neurons with two sites of synaptic integration learn invariant representations.

PubMed

Körding, K P; König, P

2001-12-01

Neurons in mammalian cerebral cortex combine specific responses with respect to some stimulus features with invariant responses to other stimulus features. For example, in primary visual cortex, complex cells code for orientation of a contour but ignore its position to a certain degree. In higher areas, such as the inferotemporal cortex, translation-invariant, rotation-invariant, and even view point-invariant responses can be observed. Such properties are of obvious interest to artificial systems performing tasks like pattern recognition. It remains to be resolved how such response properties develop in biological systems. Here we present an unsupervised learning rule that addresses this problem. It is based on a neuron model with two sites of synaptic integration, allowing qualitatively different effects of input to basal and apical dendritic trees, respectively. Without supervision, the system learns to extract invariance properties using temporal or spatial continuity of stimuli. Furthermore, top-down information can be smoothly integrated in the same framework. Thus, this model lends a physiological implementation to approaches of unsupervised learning of invariant-response properties.

Noise-robust unsupervised spike sorting based on discriminative subspace learning with outlier handling.

PubMed

Keshtkaran, Mohammad Reza; Yang, Zhi

2017-06-01

Spike sorting is a fundamental preprocessing step for many neuroscience studies which rely on the analysis of spike trains. Most of the feature extraction and dimensionality reduction techniques that have been used for spike sorting give a projection subspace which is not necessarily the most discriminative one. Therefore, the clusters which appear inherently separable in some discriminative subspace may overlap if projected using conventional feature extraction approaches leading to a poor sorting accuracy especially when the noise level is high. In this paper, we propose a noise-robust and unsupervised spike sorting algorithm based on learning discriminative spike features for clustering. The proposed algorithm uses discriminative subspace learning to extract low dimensional and most discriminative features from the spike waveforms and perform clustering with automatic detection of the number of the clusters. The core part of the algorithm involves iterative subspace selection using linear discriminant analysis and clustering using Gaussian mixture model with outlier detection. A statistical test in the discriminative subspace is proposed to automatically detect the number of the clusters. Comparative results on publicly available simulated and real in vivo datasets demonstrate that our algorithm achieves substantially improved cluster distinction leading to higher sorting accuracy and more reliable detection of clusters which are highly overlapping and not detectable using conventional feature extraction techniques such as principal component analysis or wavelets. By providing more accurate information about the activity of more number of individual neurons with high robustness to neural noise and outliers, the proposed unsupervised spike sorting algorithm facilitates more detailed and accurate analysis of single- and multi-unit activities in neuroscience and brain machine interface studies.

Noise-robust unsupervised spike sorting based on discriminative subspace learning with outlier handling

NASA Astrophysics Data System (ADS)

Keshtkaran, Mohammad Reza; Yang, Zhi

2017-06-01

Objective. Spike sorting is a fundamental preprocessing step for many neuroscience studies which rely on the analysis of spike trains. Most of the feature extraction and dimensionality reduction techniques that have been used for spike sorting give a projection subspace which is not necessarily the most discriminative one. Therefore, the clusters which appear inherently separable in some discriminative subspace may overlap if projected using conventional feature extraction approaches leading to a poor sorting accuracy especially when the noise level is high. In this paper, we propose a noise-robust and unsupervised spike sorting algorithm based on learning discriminative spike features for clustering. Approach. The proposed algorithm uses discriminative subspace learning to extract low dimensional and most discriminative features from the spike waveforms and perform clustering with automatic detection of the number of the clusters. The core part of the algorithm involves iterative subspace selection using linear discriminant analysis and clustering using Gaussian mixture model with outlier detection. A statistical test in the discriminative subspace is proposed to automatically detect the number of the clusters. Main results. Comparative results on publicly available simulated and real in vivo datasets demonstrate that our algorithm achieves substantially improved cluster distinction leading to higher sorting accuracy and more reliable detection of clusters which are highly overlapping and not detectable using conventional feature extraction techniques such as principal component analysis or wavelets. Significance. By providing more accurate information about the activity of more number of individual neurons with high robustness to neural noise and outliers, the proposed unsupervised spike sorting algorithm facilitates more detailed and accurate analysis of single- and multi-unit activities in neuroscience and brain machine interface studies.

Biologically Inspired Model for Visual Cognition Achieving Unsupervised Episodic and Semantic Feature Learning.

PubMed

Qiao, Hong; Li, Yinlin; Li, Fengfu; Xi, Xuanyang; Wu, Wei

2016-10-01

Recently, many biologically inspired visual computational models have been proposed. The design of these models follows the related biological mechanisms and structures, and these models provide new solutions for visual recognition tasks. In this paper, based on the recent biological evidence, we propose a framework to mimic the active and dynamic learning and recognition process of the primate visual cortex. From principle point of view, the main contributions are that the framework can achieve unsupervised learning of episodic features (including key components and their spatial relations) and semantic features (semantic descriptions of the key components), which support higher level cognition of an object. From performance point of view, the advantages of the framework are as follows: 1) learning episodic features without supervision-for a class of objects without a prior knowledge, the key components, their spatial relations and cover regions can be learned automatically through a deep neural network (DNN); 2) learning semantic features based on episodic features-within the cover regions of the key components, the semantic geometrical values of these components can be computed based on contour detection; 3) forming the general knowledge of a class of objects-the general knowledge of a class of objects can be formed, mainly including the key components, their spatial relations and average semantic values, which is a concise description of the class; and 4) achieving higher level cognition and dynamic updating-for a test image, the model can achieve classification and subclass semantic descriptions. And the test samples with high confidence are selected to dynamically update the whole model. Experiments are conducted on face images, and a good performance is achieved in each layer of the DNN and the semantic description learning process. Furthermore, the model can be generalized to recognition tasks of other objects with learning ability.

Learning relevant features of data with multi-scale tensor networks

NASA Astrophysics Data System (ADS)

Miles Stoudenmire, E.

2018-07-01

Inspired by coarse-graining approaches used in physics, we show how similar algorithms can be adapted for data. The resulting algorithms are based on layered tree tensor networks and scale linearly with both the dimension of the input and the training set size. Computing most of the layers with an unsupervised algorithm, then optimizing just the top layer for supervised classification of the MNIST and fashion MNIST data sets gives very good results. We also discuss mixing a prior guess for supervised weights together with an unsupervised representation of the data, yielding a smaller number of features nevertheless able to give good performance.

FRaC: a feature-modeling approach for semi-supervised and unsupervised anomaly detection.

PubMed

Noto, Keith; Brodley, Carla; Slonim, Donna

2012-01-01

Anomaly detection involves identifying rare data instances (anomalies) that come from a different class or distribution than the majority (which are simply called "normal" instances). Given a training set of only normal data, the semi-supervised anomaly detection task is to identify anomalies in the future. Good solutions to this task have applications in fraud and intrusion detection. The unsupervised anomaly detection task is different: Given unlabeled, mostly-normal data, identify the anomalies among them. Many real-world machine learning tasks, including many fraud and intrusion detection tasks, are unsupervised because it is impractical (or impossible) to verify all of the training data. We recently presented FRaC, a new approach for semi-supervised anomaly detection. FRaC is based on using normal instances to build an ensemble of feature models, and then identifying instances that disagree with those models as anomalous. In this paper, we investigate the behavior of FRaC experimentally and explain why FRaC is so successful. We also show that FRaC is a superior approach for the unsupervised as well as the semi-supervised anomaly detection task, compared to well-known state-of-the-art anomaly detection methods, LOF and one-class support vector machines, and to an existing feature-modeling approach.

FRaC: a feature-modeling approach for semi-supervised and unsupervised anomaly detection

PubMed Central

Brodley, Carla; Slonim, Donna

2011-01-01

Anomaly detection involves identifying rare data instances (anomalies) that come from a different class or distribution than the majority (which are simply called “normal” instances). Given a training set of only normal data, the semi-supervised anomaly detection task is to identify anomalies in the future. Good solutions to this task have applications in fraud and intrusion detection. The unsupervised anomaly detection task is different: Given unlabeled, mostly-normal data, identify the anomalies among them. Many real-world machine learning tasks, including many fraud and intrusion detection tasks, are unsupervised because it is impractical (or impossible) to verify all of the training data. We recently presented FRaC, a new approach for semi-supervised anomaly detection. FRaC is based on using normal instances to build an ensemble of feature models, and then identifying instances that disagree with those models as anomalous. In this paper, we investigate the behavior of FRaC experimentally and explain why FRaC is so successful. We also show that FRaC is a superior approach for the unsupervised as well as the semi-supervised anomaly detection task, compared to well-known state-of-the-art anomaly detection methods, LOF and one-class support vector machines, and to an existing feature-modeling approach. PMID:22639542

Unsupervised and self-mapping category formation and semantic object recognition for mobile robot vision used in an actual environment

NASA Astrophysics Data System (ADS)

Madokoro, H.; Tsukada, M.; Sato, K.

2013-07-01

This paper presents an unsupervised learning-based object category formation and recognition method for mobile robot vision. Our method has the following features: detection of feature points and description of features using a scale-invariant feature transform (SIFT), selection of target feature points using one class support vector machines (OC-SVMs), generation of visual words using self-organizing maps (SOMs), formation of labels using adaptive resonance theory 2 (ART-2), and creation and classification of categories on a category map of counter propagation networks (CPNs) for visualizing spatial relations between categories. Classification results of dynamic images using time-series images obtained using two different-size robots and according to movements respectively demonstrate that our method can visualize spatial relations of categories while maintaining time-series characteristics. Moreover, we emphasize the effectiveness of our method for category formation of appearance changes of objects.

Land cover classification in multispectral imagery using clustering of sparse approximations over learned feature dictionaries

DOE PAGES

Moody, Daniela I.; Brumby, Steven P.; Rowland, Joel C.; ...

2014-12-09

We present results from an ongoing effort to extend neuromimetic machine vision algorithms to multispectral data using adaptive signal processing combined with compressive sensing and machine learning techniques. Our goal is to develop a robust classification methodology that will allow for automated discretization of the landscape into distinct units based on attributes such as vegetation, surface hydrological properties, and topographic/geomorphic characteristics. We use a Hebbian learning rule to build spectral-textural dictionaries that are tailored for classification. We learn our dictionaries from millions of overlapping multispectral image patches and then use a pursuit search to generate classification features. Land cover labelsmore » are automatically generated using unsupervised clustering of sparse approximations (CoSA). We demonstrate our method on multispectral WorldView-2 data from a coastal plain ecosystem in Barrow, Alaska. We explore learning from both raw multispectral imagery and normalized band difference indices. We explore a quantitative metric to evaluate the spectral properties of the clusters in order to potentially aid in assigning land cover categories to the cluster labels. In this study, our results suggest CoSA is a promising approach to unsupervised land cover classification in high-resolution satellite imagery.« less

Image fusion using sparse overcomplete feature dictionaries

DOEpatents

Brumby, Steven P.; Bettencourt, Luis; Kenyon, Garrett T.; Chartrand, Rick; Wohlberg, Brendt

2015-10-06

Approaches for deciding what individuals in a population of visual system "neurons" are looking for using sparse overcomplete feature dictionaries are provided. A sparse overcomplete feature dictionary may be learned for an image dataset and a local sparse representation of the image dataset may be built using the learned feature dictionary. A local maximum pooling operation may be applied on the local sparse representation to produce a translation-tolerant representation of the image dataset. An object may then be classified and/or clustered within the translation-tolerant representation of the image dataset using a supervised classification algorithm and/or an unsupervised clustering algorithm.

Spatiotemporal information during unsupervised learning enhances viewpoint invariant object recognition

PubMed Central

Tian, Moqian; Grill-Spector, Kalanit

2015-01-01

Recognizing objects is difficult because it requires both linking views of an object that can be different and distinguishing objects with similar appearance. Interestingly, people can learn to recognize objects across views in an unsupervised way, without feedback, just from the natural viewing statistics. However, there is intense debate regarding what information during unsupervised learning is used to link among object views. Specifically, researchers argue whether temporal proximity, motion, or spatiotemporal continuity among object views during unsupervised learning is beneficial. Here, we untangled the role of each of these factors in unsupervised learning of novel three-dimensional (3-D) objects. We found that after unsupervised training with 24 object views spanning a 180° view space, participants showed significant improvement in their ability to recognize 3-D objects across rotation. Surprisingly, there was no advantage to unsupervised learning with spatiotemporal continuity or motion information than training with temporal proximity. However, we discovered that when participants were trained with just a third of the views spanning the same view space, unsupervised learning via spatiotemporal continuity yielded significantly better recognition performance on novel views than learning via temporal proximity. These results suggest that while it is possible to obtain view-invariant recognition just from observing many views of an object presented in temporal proximity, spatiotemporal information enhances performance by producing representations with broader view tuning than learning via temporal association. Our findings have important implications for theories of object recognition and for the development of computational algorithms that learn from examples. PMID:26024454

Weighted Distance Functions Improve Analysis of High-Dimensional Data: Application to Molecular Dynamics Simulations.

PubMed

Blöchliger, Nicolas; Caflisch, Amedeo; Vitalis, Andreas

2015-11-10

Data mining techniques depend strongly on how the data are represented and how distance between samples is measured. High-dimensional data often contain a large number of irrelevant dimensions (features) for a given query. These features act as noise and obfuscate relevant information. Unsupervised approaches to mine such data require distance measures that can account for feature relevance. Molecular dynamics simulations produce high-dimensional data sets describing molecules observed in time. Here, we propose to globally or locally weight simulation features based on effective rates. This emphasizes, in a data-driven manner, slow degrees of freedom that often report on the metastable states sampled by the molecular system. We couple this idea to several unsupervised learning protocols. Our approach unmasks slow side chain dynamics within the native state of a miniprotein and reveals additional metastable conformations of a protein. The approach can be combined with most algorithms for clustering or dimensionality reduction.

Learning Long-Range Vision for an Offroad Robot

DTIC Science & Technology

2008-09-01

robot to perceive and navigate in an unstructured natural world is a difficult task. Without learning, navigation systems are short-range and extremely...unsupervised or weakly supervised learning methods are necessary for training general feature representations for natural scenes. The process was...the world looked dark, and Legos when I was weary. iii ABSTRACT Teaching a robot to perceive and navigate in an unstructured natural world is a

A Deep Convolutional Coupling Network for Change Detection Based on Heterogeneous Optical and Radar Images.

PubMed

Liu, Jia; Gong, Maoguo; Qin, Kai; Zhang, Puzhao

2018-03-01

We propose an unsupervised deep convolutional coupling network for change detection based on two heterogeneous images acquired by optical sensors and radars on different dates. Most existing change detection methods are based on homogeneous images. Due to the complementary properties of optical and radar sensors, there is an increasing interest in change detection based on heterogeneous images. The proposed network is symmetric with each side consisting of one convolutional layer and several coupling layers. The two input images connected with the two sides of the network, respectively, are transformed into a feature space where their feature representations become more consistent. In this feature space, the different map is calculated, which then leads to the ultimate detection map by applying a thresholding algorithm. The network parameters are learned by optimizing a coupling function. The learning process is unsupervised, which is different from most existing change detection methods based on heterogeneous images. Experimental results on both homogenous and heterogeneous images demonstrate the promising performance of the proposed network compared with several existing approaches.

Unsupervised Learning of Overlapping Image Components Using Divisive Input Modulation

PubMed Central

Spratling, M. W.; De Meyer, K.; Kompass, R.

2009-01-01

This paper demonstrates that nonnegative matrix factorisation is mathematically related to a class of neural networks that employ negative feedback as a mechanism of competition. This observation inspires a novel learning algorithm which we call Divisive Input Modulation (DIM). The proposed algorithm provides a mathematically simple and computationally efficient method for the unsupervised learning of image components, even in conditions where these elementary features overlap considerably. To test the proposed algorithm, a novel artificial task is introduced which is similar to the frequently-used bars problem but employs squares rather than bars to increase the degree of overlap between components. Using this task, we investigate how the proposed method performs on the parsing of artificial images composed of overlapping features, given the correct representation of the individual components; and secondly, we investigate how well it can learn the elementary components from artificial training images. We compare the performance of the proposed algorithm with its predecessors including variations on these algorithms that have produced state-of-the-art performance on the bars problem. The proposed algorithm is more successful than its predecessors in dealing with overlap and occlusion in the artificial task that has been used to assess performance. PMID:19424442

Mere exposure alters category learning of novel objects.

PubMed

Folstein, Jonathan R; Gauthier, Isabel; Palmeri, Thomas J

2010-01-01

We investigated how mere exposure to complex objects with correlated or uncorrelated object features affects later category learning of new objects not seen during exposure. Correlations among pre-exposed object dimensions influenced later category learning. Unlike other published studies, the collection of pre-exposed objects provided no information regarding the categories to be learned, ruling out unsupervised or incidental category learning during pre-exposure. Instead, results are interpreted with respect to statistical learning mechanisms, providing one of the first demonstrations of how statistical learning can influence visual object learning.

Mere Exposure Alters Category Learning of Novel Objects

PubMed Central

Folstein, Jonathan R.; Gauthier, Isabel; Palmeri, Thomas J.

2010-01-01

We investigated how mere exposure to complex objects with correlated or uncorrelated object features affects later category learning of new objects not seen during exposure. Correlations among pre-exposed object dimensions influenced later category learning. Unlike other published studies, the collection of pre-exposed objects provided no information regarding the categories to be learned, ruling out unsupervised or incidental category learning during pre-exposure. Instead, results are interpreted with respect to statistical learning mechanisms, providing one of the first demonstrations of how statistical learning can influence visual object learning. PMID:21833209

Impact of feature saliency on visual category learning.

PubMed

Hammer, Rubi

2015-01-01

People have to sort numerous objects into a large number of meaningful categories while operating in varying contexts. This requires identifying the visual features that best predict the 'essence' of objects (e.g., edibility), rather than categorizing objects based on the most salient features in a given context. To gain this capacity, visual category learning (VCL) relies on multiple cognitive processes. These may include unsupervised statistical learning, that requires observing multiple objects for learning the statistics of their features. Other learning processes enable incorporating different sources of supervisory information, alongside the visual features of the categorized objects, from which the categorical relations between few objects can be deduced. These deductions enable inferring that objects from the same category may differ from one another in some high-saliency feature dimensions, whereas lower-saliency feature dimensions can best differentiate objects from distinct categories. Here I illustrate how feature saliency affects VCL, by also discussing kinds of supervisory information enabling reflective categorization. Arguably, principles debated here are often being ignored in categorization studies.

Impact of feature saliency on visual category learning

PubMed Central

Hammer, Rubi

2015-01-01

People have to sort numerous objects into a large number of meaningful categories while operating in varying contexts. This requires identifying the visual features that best predict the ‘essence’ of objects (e.g., edibility), rather than categorizing objects based on the most salient features in a given context. To gain this capacity, visual category learning (VCL) relies on multiple cognitive processes. These may include unsupervised statistical learning, that requires observing multiple objects for learning the statistics of their features. Other learning processes enable incorporating different sources of supervisory information, alongside the visual features of the categorized objects, from which the categorical relations between few objects can be deduced. These deductions enable inferring that objects from the same category may differ from one another in some high-saliency feature dimensions, whereas lower-saliency feature dimensions can best differentiate objects from distinct categories. Here I illustrate how feature saliency affects VCL, by also discussing kinds of supervisory information enabling reflective categorization. Arguably, principles debated here are often being ignored in categorization studies. PMID:25954220

Identifying product order with restricted Boltzmann machines

NASA Astrophysics Data System (ADS)

Rao, Wen-Jia; Li, Zhenyu; Zhu, Qiong; Luo, Mingxing; Wan, Xin

2018-03-01

Unsupervised machine learning via a restricted Boltzmann machine is a useful tool in distinguishing an ordered phase from a disordered phase. Here we study its application on the two-dimensional Ashkin-Teller model, which features a partially ordered product phase. We train the neural network with spin configuration data generated by Monte Carlo simulations and show that distinct features of the product phase can be learned from nonergodic samples resulting from symmetry breaking. Careful analysis of the weight matrices inspires us to define a nontrivial machine-learning motivated quantity of the product form, which resembles the conventional product order parameter.

Unsupervised Deep Learning Applied to Breast Density Segmentation and Mammographic Risk Scoring.

PubMed

Kallenberg, Michiel; Petersen, Kersten; Nielsen, Mads; Ng, Andrew Y; Pengfei Diao; Igel, Christian; Vachon, Celine M; Holland, Katharina; Winkel, Rikke Rass; Karssemeijer, Nico; Lillholm, Martin

2016-05-01

Mammographic risk scoring has commonly been automated by extracting a set of handcrafted features from mammograms, and relating the responses directly or indirectly to breast cancer risk. We present a method that learns a feature hierarchy from unlabeled data. When the learned features are used as the input to a simple classifier, two different tasks can be addressed: i) breast density segmentation, and ii) scoring of mammographic texture. The proposed model learns features at multiple scales. To control the models capacity a novel sparsity regularizer is introduced that incorporates both lifetime and population sparsity. We evaluated our method on three different clinical datasets. Our state-of-the-art results show that the learned breast density scores have a very strong positive relationship with manual ones, and that the learned texture scores are predictive of breast cancer. The model is easy to apply and generalizes to many other segmentation and scoring problems.

Hybrid image representation learning model with invariant features for basal cell carcinoma detection

NASA Astrophysics Data System (ADS)

Arevalo, John; Cruz-Roa, Angel; González, Fabio A.

2013-11-01

This paper presents a novel method for basal-cell carcinoma detection, which combines state-of-the-art methods for unsupervised feature learning (UFL) and bag of features (BOF) representation. BOF, which is a form of representation learning, has shown a good performance in automatic histopathology image classi cation. In BOF, patches are usually represented using descriptors such as SIFT and DCT. We propose to use UFL to learn the patch representation itself. This is accomplished by applying a topographic UFL method (T-RICA), which automatically learns visual invariance properties of color, scale and rotation from an image collection. These learned features also reveals these visual properties associated to cancerous and healthy tissues and improves carcinoma detection results by 7% with respect to traditional autoencoders, and 6% with respect to standard DCT representations obtaining in average 92% in terms of F-score and 93% of balanced accuracy.

Radio Model-free Noise Reduction of Radio Transmissions with Convolutional Autoencoders

DTIC Science & Technology

2016-09-01

Encoder-Decoder Architecture for Image Segmentation .” Cornell University Library. Computing Research Repository (CoRR). abs/1511.00561. 2. Anthony J. Bell...Aaron C Courville, and Pascal Vincent. 2012. “Unsupervised Feature Learning and Deep Learning : A Review and New Perspectives.” Cornell University...Linux Journal 122(June):1–4. 5. Francois Chollet. 2015.“Keras: Deep Learning Library for TensorFlow and Theano.” Available online at https://github.com

An introduction to kernel-based learning algorithms.

PubMed

Müller, K R; Mika, S; Rätsch, G; Tsuda, K; Schölkopf, B

2001-01-01

This paper provides an introduction to support vector machines, kernel Fisher discriminant analysis, and kernel principal component analysis, as examples for successful kernel-based learning methods. We first give a short background about Vapnik-Chervonenkis theory and kernel feature spaces and then proceed to kernel based learning in supervised and unsupervised scenarios including practical and algorithmic considerations. We illustrate the usefulness of kernel algorithms by discussing applications such as optical character recognition and DNA analysis.

Robust Joint Graph Sparse Coding for Unsupervised Spectral Feature Selection.

PubMed

Zhu, Xiaofeng; Li, Xuelong; Zhang, Shichao; Ju, Chunhua; Wu, Xindong

2017-06-01

In this paper, we propose a new unsupervised spectral feature selection model by embedding a graph regularizer into the framework of joint sparse regression for preserving the local structures of data. To do this, we first extract the bases of training data by previous dictionary learning methods and, then, map original data into the basis space to generate their new representations, by proposing a novel joint graph sparse coding (JGSC) model. In JGSC, we first formulate its objective function by simultaneously taking subspace learning and joint sparse regression into account, then, design a new optimization solution to solve the resulting objective function, and further prove the convergence of the proposed solution. Furthermore, we extend JGSC to a robust JGSC (RJGSC) via replacing the least square loss function with a robust loss function, for achieving the same goals and also avoiding the impact of outliers. Finally, experimental results on real data sets showed that both JGSC and RJGSC outperformed the state-of-the-art algorithms in terms of k -nearest neighbor classification performance.

Simultaneous Local Binary Feature Learning and Encoding for Homogeneous and Heterogeneous Face Recognition.

PubMed

Lu, Jiwen; Erin Liong, Venice; Zhou, Jie

2017-08-09

In this paper, we propose a simultaneous local binary feature learning and encoding (SLBFLE) approach for both homogeneous and heterogeneous face recognition. Unlike existing hand-crafted face descriptors such as local binary pattern (LBP) and Gabor features which usually require strong prior knowledge, our SLBFLE is an unsupervised feature learning approach which automatically learns face representation from raw pixels. Unlike existing binary face descriptors such as the LBP, discriminant face descriptor (DFD), and compact binary face descriptor (CBFD) which use a two-stage feature extraction procedure, our SLBFLE jointly learns binary codes and the codebook for local face patches so that discriminative information from raw pixels from face images of different identities can be obtained by using a one-stage feature learning and encoding procedure. Moreover, we propose a coupled simultaneous local binary feature learning and encoding (C-SLBFLE) method to make the proposed approach suitable for heterogeneous face matching. Unlike most existing coupled feature learning methods which learn a pair of transformation matrices for each modality, we exploit both the common and specific information from heterogeneous face samples to characterize their underlying correlations. Experimental results on six widely used face datasets are presented to demonstrate the effectiveness of the proposed method.

Change detection and change monitoring of natural and man-made features in multispectral and hyperspectral satellite imagery

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

Moody, Daniela Irina

An approach for land cover classification, seasonal and yearly change detection and monitoring, and identification of changes in man-made features may use a clustering of sparse approximations (CoSA) on sparse representations in learned dictionaries. A Hebbian learning rule may be used to build multispectral or hyperspectral, multiresolution dictionaries that are adapted to regional satellite image data. Sparse image representations of pixel patches over the learned dictionaries may be used to perform unsupervised k-means clustering into land cover categories. The clustering process behaves as a classifier in detecting real variability. This approach may combine spectral and spatial textural characteristics to detectmore » geologic, vegetative, hydrologic, and man-made features, as well as changes in these features over time.« less

The Livermore Brain: Massive Deep Learning Networks Enabled by High Performance Computing

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

Chen, Barry Y.

The proliferation of inexpensive sensor technologies like the ubiquitous digital image sensors has resulted in the collection and sharing of vast amounts of unsorted and unexploited raw data. Companies and governments who are able to collect and make sense of large datasets to help them make better decisions more rapidly will have a competitive advantage in the information era. Machine Learning technologies play a critical role for automating the data understanding process; however, to be maximally effective, useful intermediate representations of the data are required. These representations or “features” are transformations of the raw data into a form where patternsmore » are more easily recognized. Recent breakthroughs in Deep Learning have made it possible to learn these features from large amounts of labeled data. The focus of this project is to develop and extend Deep Learning algorithms for learning features from vast amounts of unlabeled data and to develop the HPC neural network training platform to support the training of massive network models. This LDRD project succeeded in developing new unsupervised feature learning algorithms for images and video and created a scalable neural network training toolkit for HPC. Additionally, this LDRD helped create the world’s largest freely-available image and video dataset supporting open multimedia research and used this dataset for training our deep neural networks. This research helped LLNL capture several work-for-others (WFO) projects, attract new talent, and establish collaborations with leading academic and commercial partners. Finally, this project demonstrated the successful training of the largest unsupervised image neural network using HPC resources and helped establish LLNL leadership at the intersection of Machine Learning and HPC research.« less

Automated age-related macular degeneration classification in OCT using unsupervised feature learning

NASA Astrophysics Data System (ADS)

Venhuizen, Freerk G.; van Ginneken, Bram; Bloemen, Bart; van Grinsven, Mark J. J. P.; Philipsen, Rick; Hoyng, Carel; Theelen, Thomas; Sánchez, Clara I.

2015-03-01

Age-related Macular Degeneration (AMD) is a common eye disorder with high prevalence in elderly people. The disease mainly affects the central part of the retina, and could ultimately lead to permanent vision loss. Optical Coherence Tomography (OCT) is becoming the standard imaging modality in diagnosis of AMD and the assessment of its progression. However, the evaluation of the obtained volumetric scan is time consuming, expensive and the signs of early AMD are easy to miss. In this paper we propose a classification method to automatically distinguish AMD patients from healthy subjects with high accuracy. The method is based on an unsupervised feature learning approach, and processes the complete image without the need for an accurate pre-segmentation of the retina. The method can be divided in two steps: an unsupervised clustering stage that extracts a set of small descriptive image patches from the training data, and a supervised training stage that uses these patches to create a patch occurrence histogram for every image on which a random forest classifier is trained. Experiments using 384 volume scans show that the proposed method is capable of identifying AMD patients with high accuracy, obtaining an area under the Receiver Operating Curve of 0:984. Our method allows for a quick and reliable assessment of the presence of AMD pathology in OCT volume scans without the need for accurate layer segmentation algorithms.

Transformer fault diagnosis using continuous sparse autoencoder.

PubMed

Wang, Lukun; Zhao, Xiaoying; Pei, Jiangnan; Tang, Gongyou

2016-01-01

This paper proposes a novel continuous sparse autoencoder (CSAE) which can be used in unsupervised feature learning. The CSAE adds Gaussian stochastic unit into activation function to extract features of nonlinear data. In this paper, CSAE is applied to solve the problem of transformer fault recognition. Firstly, based on dissolved gas analysis method, IEC three ratios are calculated by the concentrations of dissolved gases. Then IEC three ratios data is normalized to reduce data singularity and improve training speed. Secondly, deep belief network is established by two layers of CSAE and one layer of back propagation (BP) network. Thirdly, CSAE is adopted to unsupervised training and getting features. Then BP network is used for supervised training and getting transformer fault. Finally, the experimental data from IEC TC 10 dataset aims to illustrate the effectiveness of the presented approach. Comparative experiments clearly show that CSAE can extract features from the original data, and achieve a superior correct differentiation rate on transformer fault diagnosis.

Semi-Supervised Clustering for High-Dimensional and Sparse Features

ERIC Educational Resources Information Center

Yan, Su

2010-01-01

Clustering is one of the most common data mining tasks, used frequently for data organization and analysis in various application domains. Traditional machine learning approaches to clustering are fully automated and unsupervised where class labels are unknown a priori. In real application domains, however, some "weak" form of side…

Semi-automated surface mapping via unsupervised classification

NASA Astrophysics Data System (ADS)

D'Amore, M.; Le Scaon, R.; Helbert, J.; Maturilli, A.

2017-09-01

Due to the increasing volume of the returned data from space mission, the human search for correlation and identification of interesting features becomes more and more unfeasible. Statistical extraction of features via machine learning methods will increase the scientific output of remote sensing missions and aid the discovery of yet unknown feature hidden in dataset. Those methods exploit algorithm trained on features from multiple instrument, returning classification maps that explore intra-dataset correlation, allowing for the discovery of unknown features. We present two applications, one for Mercury and one for Vesta.

A recurrent neural network for classification of unevenly sampled variable stars

NASA Astrophysics Data System (ADS)

Naul, Brett; Bloom, Joshua S.; Pérez, Fernando; van der Walt, Stéfan

2018-02-01

Astronomical surveys of celestial sources produce streams of noisy time series measuring flux versus time (`light curves'). Unlike in many other physical domains, however, large (and source-specific) temporal gaps in data arise naturally due to intranight cadence choices as well as diurnal and seasonal constraints1-5. With nightly observations of millions of variable stars and transients from upcoming surveys4,6, efficient and accurate discovery and classification techniques on noisy, irregularly sampled data must be employed with minimal human-in-the-loop involvement. Machine learning for inference tasks on such data traditionally requires the laborious hand-coding of domain-specific numerical summaries of raw data (`features')7. Here, we present a novel unsupervised autoencoding recurrent neural network8 that makes explicit use of sampling times and known heteroskedastic noise properties. When trained on optical variable star catalogues, this network produces supervised classification models that rival other best-in-class approaches. We find that autoencoded features learned in one time-domain survey perform nearly as well when applied to another survey. These networks can continue to learn from new unlabelled observations and may be used in other unsupervised tasks, such as forecasting and anomaly detection.

Binary Multidimensional Scaling for Hashing.

PubMed

Huang, Yameng; Lin, Zhouchen

2017-10-04

Hashing is a useful technique for fast nearest neighbor search due to its low storage cost and fast query speed. Unsupervised hashing aims at learning binary hash codes for the original features so that the pairwise distances can be best preserved. While several works have targeted on this task, the results are not satisfactory mainly due to the oversimplified model. In this paper, we propose a unified and concise unsupervised hashing framework, called Binary Multidimensional Scaling (BMDS), which is able to learn the hash code for distance preservation in both batch and online mode. In the batch mode, unlike most existing hashing methods, we do not need to simplify the model by predefining the form of hash map. Instead, we learn the binary codes directly based on the pairwise distances among the normalized original features by Alternating Minimization. This enables a stronger expressive power of the hash map. In the online mode, we consider the holistic distance relationship between current query example and those we have already learned, rather than only focusing on current data chunk. It is useful when the data come in a streaming fashion. Empirical results show that while being efficient for training, our algorithm outperforms state-of-the-art methods by a large margin in terms of distance preservation, which is practical for real-world applications.

Deep Patient: An Unsupervised Representation to Predict the Future of Patients from the Electronic Health Records

PubMed Central

Miotto, Riccardo; Li, Li; Kidd, Brian A.; Dudley, Joel T.

2016-01-01

Secondary use of electronic health records (EHRs) promises to advance clinical research and better inform clinical decision making. Challenges in summarizing and representing patient data prevent widespread practice of predictive modeling using EHRs. Here we present a novel unsupervised deep feature learning method to derive a general-purpose patient representation from EHR data that facilitates clinical predictive modeling. In particular, a three-layer stack of denoising autoencoders was used to capture hierarchical regularities and dependencies in the aggregated EHRs of about 700,000 patients from the Mount Sinai data warehouse. The result is a representation we name “deep patient”. We evaluated this representation as broadly predictive of health states by assessing the probability of patients to develop various diseases. We performed evaluation using 76,214 test patients comprising 78 diseases from diverse clinical domains and temporal windows. Our results significantly outperformed those achieved using representations based on raw EHR data and alternative feature learning strategies. Prediction performance for severe diabetes, schizophrenia, and various cancers were among the top performing. These findings indicate that deep learning applied to EHRs can derive patient representations that offer improved clinical predictions, and could provide a machine learning framework for augmenting clinical decision systems. PMID:27185194

Deep Patient: An Unsupervised Representation to Predict the Future of Patients from the Electronic Health Records

NASA Astrophysics Data System (ADS)

Miotto, Riccardo; Li, Li; Kidd, Brian A.; Dudley, Joel T.

2016-05-01

Secondary use of electronic health records (EHRs) promises to advance clinical research and better inform clinical decision making. Challenges in summarizing and representing patient data prevent widespread practice of predictive modeling using EHRs. Here we present a novel unsupervised deep feature learning method to derive a general-purpose patient representation from EHR data that facilitates clinical predictive modeling. In particular, a three-layer stack of denoising autoencoders was used to capture hierarchical regularities and dependencies in the aggregated EHRs of about 700,000 patients from the Mount Sinai data warehouse. The result is a representation we name “deep patient”. We evaluated this representation as broadly predictive of health states by assessing the probability of patients to develop various diseases. We performed evaluation using 76,214 test patients comprising 78 diseases from diverse clinical domains and temporal windows. Our results significantly outperformed those achieved using representations based on raw EHR data and alternative feature learning strategies. Prediction performance for severe diabetes, schizophrenia, and various cancers were among the top performing. These findings indicate that deep learning applied to EHRs can derive patient representations that offer improved clinical predictions, and could provide a machine learning framework for augmenting clinical decision systems.

Self-supervised ARTMAP.

PubMed

Amis, Gregory P; Carpenter, Gail A

2010-03-01

Computational models of learning typically train on labeled input patterns (supervised learning), unlabeled input patterns (unsupervised learning), or a combination of the two (semi-supervised learning). In each case input patterns have a fixed number of features throughout training and testing. Human and machine learning contexts present additional opportunities for expanding incomplete knowledge from formal training, via self-directed learning that incorporates features not previously experienced. This article defines a new self-supervised learning paradigm to address these richer learning contexts, introducing a neural network called self-supervised ARTMAP. Self-supervised learning integrates knowledge from a teacher (labeled patterns with some features), knowledge from the environment (unlabeled patterns with more features), and knowledge from internal model activation (self-labeled patterns). Self-supervised ARTMAP learns about novel features from unlabeled patterns without destroying partial knowledge previously acquired from labeled patterns. A category selection function bases system predictions on known features, and distributed network activation scales unlabeled learning to prediction confidence. Slow distributed learning on unlabeled patterns focuses on novel features and confident predictions, defining classification boundaries that were ambiguous in the labeled patterns. Self-supervised ARTMAP improves test accuracy on illustrative low-dimensional problems and on high-dimensional benchmarks. Model code and benchmark data are available from: http://techlab.eu.edu/SSART/. Copyright 2009 Elsevier Ltd. All rights reserved.

Model-Based Learning of Local Image Features for Unsupervised Texture Segmentation

NASA Astrophysics Data System (ADS)

Kiechle, Martin; Storath, Martin; Weinmann, Andreas; Kleinsteuber, Martin

2018-04-01

Features that capture well the textural patterns of a certain class of images are crucial for the performance of texture segmentation methods. The manual selection of features or designing new ones can be a tedious task. Therefore, it is desirable to automatically adapt the features to a certain image or class of images. Typically, this requires a large set of training images with similar textures and ground truth segmentation. In this work, we propose a framework to learn features for texture segmentation when no such training data is available. The cost function for our learning process is constructed to match a commonly used segmentation model, the piecewise constant Mumford-Shah model. This means that the features are learned such that they provide an approximately piecewise constant feature image with a small jump set. Based on this idea, we develop a two-stage algorithm which first learns suitable convolutional features and then performs a segmentation. We note that the features can be learned from a small set of images, from a single image, or even from image patches. The proposed method achieves a competitive rank in the Prague texture segmentation benchmark, and it is effective for segmenting histological images.

Machine-Learned Data Structures of Lipid Marker Serum Concentrations in Multiple Sclerosis Patients Differ from Those in Healthy Subjects.

PubMed

Lötsch, Jörn; Thrun, Michael; Lerch, Florian; Brunkhorst, Robert; Schiffmann, Susanne; Thomas, Dominique; Tegder, Irmgard; Geisslinger, Gerd; Ultsch, Alfred

2017-06-07

Lipid metabolism has been suggested to be a major pathophysiological mechanism of multiple sclerosis (MS). With the increasing knowledge about lipid signaling, acquired data become increasingly complex making bioinformatics necessary in lipid research. We used unsupervised machine-learning to analyze lipid marker serum concentrations, pursuing the hypothesis that for the most relevant markers the emerging data structures will coincide with the diagnosis of MS. Machine learning was implemented as emergent self-organizing feature maps (ESOM) combined with the U*-matrix visualization technique. The data space consisted of serum concentrations of three main classes of lipid markers comprising eicosanoids ( d = 11 markers), ceramides ( d = 10), and lyosophosphatidic acids ( d = 6). They were analyzed in cohorts of MS patients ( n = 102) and healthy subjects ( n = 301). Clear data structures in the high-dimensional data space were observed in eicosanoid and ceramides serum concentrations whereas no clear structure could be found in lysophosphatidic acid concentrations. With ceramide concentrations, the structures that had emerged from unsupervised machine-learning almost completely overlapped with the known grouping of MS patients versus healthy subjects. This was only partly provided by eicosanoid serum concentrations. Thus, unsupervised machine-learning identified distinct data structures of bioactive lipid serum concentrations. These structures could be superimposed with the known grouping of MS patients versus healthy subjects, which was almost completely possible with ceramides. Therefore, based on the present analysis, ceramides are first-line candidates for further exploration as drug-gable targets or biomarkers in MS.

Classification of multispectral or hyperspectral satellite imagery using clustering of sparse approximations on sparse representations in learned dictionaries obtained using efficient convolutional sparse coding

DOEpatents

Moody, Daniela; Wohlberg, Brendt

2018-01-02

An approach for land cover classification, seasonal and yearly change detection and monitoring, and identification of changes in man-made features may use a clustering of sparse approximations (CoSA) on sparse representations in learned dictionaries. The learned dictionaries may be derived using efficient convolutional sparse coding to build multispectral or hyperspectral, multiresolution dictionaries that are adapted to regional satellite image data. Sparse image representations of images over the learned dictionaries may be used to perform unsupervised k-means clustering into land cover categories. The clustering process behaves as a classifier in detecting real variability. This approach may combine spectral and spatial textural characteristics to detect geologic, vegetative, hydrologic, and man-made features, as well as changes in these features over time.

Unsupervised Fault Diagnosis of a Gear Transmission Chain Using a Deep Belief Network

PubMed Central

He, Jun; Yang, Shixi; Gan, Chunbiao

2017-01-01

Artificial intelligence (AI) techniques, which can effectively analyze massive amounts of fault data and automatically provide accurate diagnosis results, have been widely applied to fault diagnosis of rotating machinery. Conventional AI methods are applied using features selected by a human operator, which are manually extracted based on diagnostic techniques and field expertise. However, developing robust features for each diagnostic purpose is often labour-intensive and time-consuming, and the features extracted for one specific task may be unsuitable for others. In this paper, a novel AI method based on a deep belief network (DBN) is proposed for the unsupervised fault diagnosis of a gear transmission chain, and the genetic algorithm is used to optimize the structural parameters of the network. Compared to the conventional AI methods, the proposed method can adaptively exploit robust features related to the faults by unsupervised feature learning, thus requires less prior knowledge about signal processing techniques and diagnostic expertise. Besides, it is more powerful at modelling complex structured data. The effectiveness of the proposed method is validated using datasets from rolling bearings and gearbox. To show the superiority of the proposed method, its performance is compared with two well-known classifiers, i.e., back propagation neural network (BPNN) and support vector machine (SVM). The fault classification accuracies are 99.26% for rolling bearings and 100% for gearbox when using the proposed method, which are much higher than that of the other two methods. PMID:28677638

Unsupervised Fault Diagnosis of a Gear Transmission Chain Using a Deep Belief Network.

PubMed

He, Jun; Yang, Shixi; Gan, Chunbiao

2017-07-04

Artificial intelligence (AI) techniques, which can effectively analyze massive amounts of fault data and automatically provide accurate diagnosis results, have been widely applied to fault diagnosis of rotating machinery. Conventional AI methods are applied using features selected by a human operator, which are manually extracted based on diagnostic techniques and field expertise. However, developing robust features for each diagnostic purpose is often labour-intensive and time-consuming, and the features extracted for one specific task may be unsuitable for others. In this paper, a novel AI method based on a deep belief network (DBN) is proposed for the unsupervised fault diagnosis of a gear transmission chain, and the genetic algorithm is used to optimize the structural parameters of the network. Compared to the conventional AI methods, the proposed method can adaptively exploit robust features related to the faults by unsupervised feature learning, thus requires less prior knowledge about signal processing techniques and diagnostic expertise. Besides, it is more powerful at modelling complex structured data. The effectiveness of the proposed method is validated using datasets from rolling bearings and gearbox. To show the superiority of the proposed method, its performance is compared with two well-known classifiers, i.e., back propagation neural network (BPNN) and support vector machine (SVM). The fault classification accuracies are 99.26% for rolling bearings and 100% for gearbox when using the proposed method, which are much higher than that of the other two methods.

An Evaluation of Feature Learning Methods for High Resolution Image Classification

NASA Astrophysics Data System (ADS)

Tokarczyk, P.; Montoya, J.; Schindler, K.

2012-07-01

Automatic image classification is one of the fundamental problems of remote sensing research. The classification problem is even more challenging in high-resolution images of urban areas, where the objects are small and heterogeneous. Two questions arise, namely which features to extract from the raw sensor data to capture the local radiometry and image structure at each pixel or segment, and which classification method to apply to the feature vectors. While classifiers are nowadays well understood, selecting the right features remains a largely empirical process. Here we concentrate on the features. Several methods are evaluated which allow one to learn suitable features from unlabelled image data by analysing the image statistics. In a comparative study, we evaluate unsupervised feature learning with different linear and non-linear learning methods, including principal component analysis (PCA) and deep belief networks (DBN). We also compare these automatically learned features with popular choices of ad-hoc features including raw intensity values, standard combinations like the NDVI, a few PCA channels, and texture filters. The comparison is done in a unified framework using the same images, the target classes, reference data and a Random Forest classifier.

Unsupervised universal steganalyzer for high-dimensional steganalytic features

NASA Astrophysics Data System (ADS)

Hou, Xiaodan; Zhang, Tao

2016-11-01

The research in developing steganalytic features has been highly successful. These features are extremely powerful when applied to supervised binary classification problems. However, they are incompatible with unsupervised universal steganalysis because the unsupervised method cannot distinguish embedding distortion from varying levels of noises caused by cover variation. This study attempts to alleviate the problem by introducing similarity retrieval of image statistical properties (SRISP), with the specific aim of mitigating the effect of cover variation on the existing steganalytic features. First, cover images with some statistical properties similar to those of a given test image are searched from a retrieval cover database to establish an aided sample set. Then, unsupervised outlier detection is performed on a test set composed of the given test image and its aided sample set to determine the type (cover or stego) of the given test image. Our proposed framework, called SRISP-aided unsupervised outlier detection, requires no training. Thus, it does not suffer from model mismatch mess. Compared with prior unsupervised outlier detectors that do not consider SRISP, the proposed framework not only retains the universality but also exhibits superior performance when applied to high-dimensional steganalytic features.

Unsupervised active learning based on hierarchical graph-theoretic clustering.

PubMed

Hu, Weiming; Hu, Wei; Xie, Nianhua; Maybank, Steve

2009-10-01

Most existing active learning approaches are supervised. Supervised active learning has the following problems: inefficiency in dealing with the semantic gap between the distribution of samples in the feature space and their labels, lack of ability in selecting new samples that belong to new categories that have not yet appeared in the training samples, and lack of adaptability to changes in the semantic interpretation of sample categories. To tackle these problems, we propose an unsupervised active learning framework based on hierarchical graph-theoretic clustering. In the framework, two promising graph-theoretic clustering algorithms, namely, dominant-set clustering and spectral clustering, are combined in a hierarchical fashion. Our framework has some advantages, such as ease of implementation, flexibility in architecture, and adaptability to changes in the labeling. Evaluations on data sets for network intrusion detection, image classification, and video classification have demonstrated that our active learning framework can effectively reduce the workload of manual classification while maintaining a high accuracy of automatic classification. It is shown that, overall, our framework outperforms the support-vector-machine-based supervised active learning, particularly in terms of dealing much more efficiently with new samples whose categories have not yet appeared in the training samples.

Robust Real-Time Music Transcription with a Compositional Hierarchical Model.

PubMed

Pesek, Matevž; Leonardis, Aleš; Marolt, Matija

2017-01-01

The paper presents a new compositional hierarchical model for robust music transcription. Its main features are unsupervised learning of a hierarchical representation of input data, transparency, which enables insights into the learned representation, as well as robustness and speed which make it suitable for real-world and real-time use. The model consists of multiple layers, each composed of a number of parts. The hierarchical nature of the model corresponds well to hierarchical structures in music. The parts in lower layers correspond to low-level concepts (e.g. tone partials), while the parts in higher layers combine lower-level representations into more complex concepts (tones, chords). The layers are learned in an unsupervised manner from music signals. Parts in each layer are compositions of parts from previous layers based on statistical co-occurrences as the driving force of the learning process. In the paper, we present the model's structure and compare it to other hierarchical approaches in the field of music information retrieval. We evaluate the model's performance for the multiple fundamental frequency estimation. Finally, we elaborate on extensions of the model towards other music information retrieval tasks.

Unsupervised learning of temporal features for word categorization in a spiking neural network model of the auditory brain.

PubMed

Higgins, Irina; Stringer, Simon; Schnupp, Jan

2017-01-01

The nature of the code used in the auditory cortex to represent complex auditory stimuli, such as naturally spoken words, remains a matter of debate. Here we argue that such representations are encoded by stable spatio-temporal patterns of firing within cell assemblies known as polychronous groups, or PGs. We develop a physiologically grounded, unsupervised spiking neural network model of the auditory brain with local, biologically realistic, spike-time dependent plasticity (STDP) learning, and show that the plastic cortical layers of the network develop PGs which convey substantially more information about the speaker independent identity of two naturally spoken word stimuli than does rate encoding that ignores the precise spike timings. We furthermore demonstrate that such informative PGs can only develop if the input spatio-temporal spike patterns to the plastic cortical areas of the model are relatively stable.

Metric Learning to Enhance Hyperspectral Image Segmentation

NASA Technical Reports Server (NTRS)

Thompson, David R.; Castano, Rebecca; Bue, Brian; Gilmore, Martha S.

2013-01-01

Unsupervised hyperspectral image segmentation can reveal spatial trends that show the physical structure of the scene to an analyst. They highlight borders and reveal areas of homogeneity and change. Segmentations are independently helpful for object recognition, and assist with automated production of symbolic maps. Additionally, a good segmentation can dramatically reduce the number of effective spectra in an image, enabling analyses that would otherwise be computationally prohibitive. Specifically, using an over-segmentation of the image instead of individual pixels can reduce noise and potentially improve the results of statistical post-analysis. In this innovation, a metric learning approach is presented to improve the performance of unsupervised hyperspectral image segmentation. The prototype demonstrations attempt a superpixel segmentation in which the image is conservatively over-segmented; that is, the single surface features may be split into multiple segments, but each individual segment, or superpixel, is ensured to have homogenous mineralogy.

Unsupervised learning of temporal features for word categorization in a spiking neural network model of the auditory brain

PubMed Central

Stringer, Simon

2017-01-01

The nature of the code used in the auditory cortex to represent complex auditory stimuli, such as naturally spoken words, remains a matter of debate. Here we argue that such representations are encoded by stable spatio-temporal patterns of firing within cell assemblies known as polychronous groups, or PGs. We develop a physiologically grounded, unsupervised spiking neural network model of the auditory brain with local, biologically realistic, spike-time dependent plasticity (STDP) learning, and show that the plastic cortical layers of the network develop PGs which convey substantially more information about the speaker independent identity of two naturally spoken word stimuli than does rate encoding that ignores the precise spike timings. We furthermore demonstrate that such informative PGs can only develop if the input spatio-temporal spike patterns to the plastic cortical areas of the model are relatively stable. PMID:28797034

Unsupervised Learning in an Ensemble of Spiking Neural Networks Mediated by ITDP.

PubMed

Shim, Yoonsik; Philippides, Andrew; Staras, Kevin; Husbands, Phil

2016-10-01

We propose a biologically plausible architecture for unsupervised ensemble learning in a population of spiking neural network classifiers. A mixture of experts type organisation is shown to be effective, with the individual classifier outputs combined via a gating network whose operation is driven by input timing dependent plasticity (ITDP). The ITDP gating mechanism is based on recent experimental findings. An abstract, analytically tractable model of the ITDP driven ensemble architecture is derived from a logical model based on the probabilities of neural firing events. A detailed analysis of this model provides insights that allow it to be extended into a full, biologically plausible, computational implementation of the architecture which is demonstrated on a visual classification task. The extended model makes use of a style of spiking network, first introduced as a model of cortical microcircuits, that is capable of Bayesian inference, effectively performing expectation maximization. The unsupervised ensemble learning mechanism, based around such spiking expectation maximization (SEM) networks whose combined outputs are mediated by ITDP, is shown to perform the visual classification task well and to generalize to unseen data. The combined ensemble performance is significantly better than that of the individual classifiers, validating the ensemble architecture and learning mechanisms. The properties of the full model are analysed in the light of extensive experiments with the classification task, including an investigation into the influence of different input feature selection schemes and a comparison with a hierarchical STDP based ensemble architecture.

Unsupervised Learning in an Ensemble of Spiking Neural Networks Mediated by ITDP

PubMed Central

Staras, Kevin

2016-01-01

We propose a biologically plausible architecture for unsupervised ensemble learning in a population of spiking neural network classifiers. A mixture of experts type organisation is shown to be effective, with the individual classifier outputs combined via a gating network whose operation is driven by input timing dependent plasticity (ITDP). The ITDP gating mechanism is based on recent experimental findings. An abstract, analytically tractable model of the ITDP driven ensemble architecture is derived from a logical model based on the probabilities of neural firing events. A detailed analysis of this model provides insights that allow it to be extended into a full, biologically plausible, computational implementation of the architecture which is demonstrated on a visual classification task. The extended model makes use of a style of spiking network, first introduced as a model of cortical microcircuits, that is capable of Bayesian inference, effectively performing expectation maximization. The unsupervised ensemble learning mechanism, based around such spiking expectation maximization (SEM) networks whose combined outputs are mediated by ITDP, is shown to perform the visual classification task well and to generalize to unseen data. The combined ensemble performance is significantly better than that of the individual classifiers, validating the ensemble architecture and learning mechanisms. The properties of the full model are analysed in the light of extensive experiments with the classification task, including an investigation into the influence of different input feature selection schemes and a comparison with a hierarchical STDP based ensemble architecture. PMID:27760125

Machine-Learned Data Structures of Lipid Marker Serum Concentrations in Multiple Sclerosis Patients Differ from Those in Healthy Subjects

PubMed Central

Lötsch, Jörn; Thrun, Michael; Lerch, Florian; Brunkhorst, Robert; Schiffmann, Susanne; Thomas, Dominique; Tegder, Irmgard; Geisslinger, Gerd; Ultsch, Alfred

2017-01-01

Lipid signaling has been suggested to be a major pathophysiological mechanism of multiple sclerosis (MS). With the increasing knowledge about lipid signaling, acquired data become increasingly complex making bioinformatics necessary in lipid research. We used unsupervised machine-learning to analyze lipid marker serum concentrations, pursuing the hypothesis that for the most relevant markers the emerging data structures will coincide with the diagnosis of MS. Machine learning was implemented as emergent self-organizing feature maps (ESOM) combined with the U*-matrix visualization technique. The data space consisted of serum concentrations of three main classes of lipid markers comprising eicosanoids (d = 11 markers), ceramides (d = 10), and lyosophosphatidic acids (d = 6). They were analyzed in cohorts of MS patients (n = 102) and healthy subjects (n = 301). Clear data structures in the high-dimensional data space were observed in eicosanoid and ceramides serum concentrations whereas no clear structure could be found in lysophosphatidic acid concentrations. With ceramide concentrations, the structures that had emerged from unsupervised machine-learning almost completely overlapped with the known grouping of MS patients versus healthy subjects. This was only partly provided by eicosanoid serum concentrations. Thus, unsupervised machine-learning identified distinct data structures of bioactive lipid serum concentrations. These structures could be superimposed with the known grouping of MS patients versus healthy subjects, which was almost completely possible with ceramides. Therefore, based on the present analysis, ceramides are first-line candidates for further exploration as drug-gable targets or biomarkers in MS. PMID:28590455

Unsupervised learning on scientific ocean drilling datasets from the South China Sea

NASA Astrophysics Data System (ADS)

Tse, Kevin C.; Chiu, Hon-Chim; Tsang, Man-Yin; Li, Yiliang; Lam, Edmund Y.

2018-06-01

Unsupervised learning methods were applied to explore data patterns in multivariate geophysical datasets collected from ocean floor sediment core samples coming from scientific ocean drilling in the South China Sea. Compared to studies on similar datasets, but using supervised learning methods which are designed to make predictions based on sample training data, unsupervised learning methods require no a priori information and focus only on the input data. In this study, popular unsupervised learning methods including K-means, self-organizing maps, hierarchical clustering and random forest were coupled with different distance metrics to form exploratory data clusters. The resulting data clusters were externally validated with lithologic units and geologic time scales assigned to the datasets by conventional methods. Compact and connected data clusters displayed varying degrees of correspondence with existing classification by lithologic units and geologic time scales. K-means and self-organizing maps were observed to perform better with lithologic units while random forest corresponded best with geologic time scales. This study sets a pioneering example of how unsupervised machine learning methods can be used as an automatic processing tool for the increasingly high volume of scientific ocean drilling data.

Visual texture perception via graph-based semi-supervised learning

NASA Astrophysics Data System (ADS)

Zhang, Qin; Dong, Junyu; Zhong, Guoqiang

2018-04-01

Perceptual features, for example direction, contrast and repetitiveness, are important visual factors for human to perceive a texture. However, it needs to perform psychophysical experiment to quantify these perceptual features' scale, which requires a large amount of human labor and time. This paper focuses on the task of obtaining perceptual features' scale of textures by small number of textures with perceptual scales through a rating psychophysical experiment (what we call labeled textures) and a mass of unlabeled textures. This is the scenario that the semi-supervised learning is naturally suitable for. This is meaningful for texture perception research, and really helpful for the perceptual texture database expansion. A graph-based semi-supervised learning method called random multi-graphs, RMG for short, is proposed to deal with this task. We evaluate different kinds of features including LBP, Gabor, and a kind of unsupervised deep features extracted by a PCA-based deep network. The experimental results show that our method can achieve satisfactory effects no matter what kind of texture features are used.

Two-layer contractive encodings for learning stable nonlinear features.

PubMed

Schulz, Hannes; Cho, Kyunghyun; Raiko, Tapani; Behnke, Sven

2015-04-01

Unsupervised learning of feature hierarchies is often a good strategy to initialize deep architectures for supervised learning. Most existing deep learning methods build these feature hierarchies layer by layer in a greedy fashion using either auto-encoders or restricted Boltzmann machines. Both yield encoders which compute linear projections of input followed by a smooth thresholding function. In this work, we demonstrate that these encoders fail to find stable features when the required computation is in the exclusive-or class. To overcome this limitation, we propose a two-layer encoder which is less restricted in the type of features it can learn. The proposed encoder is regularized by an extension of previous work on contractive regularization. This proposed two-layer contractive encoder potentially poses a more difficult optimization problem, and we further propose to linearly transform hidden neurons of the encoder to make learning easier. We demonstrate the advantages of the two-layer encoders qualitatively on artificially constructed datasets as well as commonly used benchmark datasets. We also conduct experiments on a semi-supervised learning task and show the benefits of the proposed two-layer encoders trained with the linear transformation of perceptrons. Copyright © 2014 Elsevier Ltd. All rights reserved.

Psoriasis image representation using patch-based dictionary learning for erythema severity scoring.

PubMed

George, Yasmeen; Aldeen, Mohammad; Garnavi, Rahil

2018-06-01

Psoriasis is a chronic skin disease which can be life-threatening. Accurate severity scoring helps dermatologists to decide on the treatment. In this paper, we present a semi-supervised computer-aided system for automatic erythema severity scoring in psoriasis images. Firstly, the unsupervised stage includes a novel image representation method. We construct a dictionary, which is then used in the sparse representation for local feature extraction. To acquire the final image representation vector, an aggregation method is exploited over the local features. Secondly, the supervised phase is where various multi-class machine learning (ML) classifiers are trained for erythema severity scoring. Finally, we compare the proposed system with two popular unsupervised feature extractor methods, namely: bag of visual words model (BoVWs) and AlexNet pretrained model. Root mean square error (RMSE) and F1 score are used as performance measures for the learned dictionaries and the trained ML models, respectively. A psoriasis image set consisting of 676 images, is used in this study. Experimental results demonstrate that the use of the proposed procedure can provide a setup where erythema scoring is accurate and consistent. Also, it is revealed that dictionaries with large number of atoms and small patch sizes yield the best representative erythema severity features. Further, random forest (RF) outperforms other classifiers with F1 score 0.71, followed by support vector machine (SVM) and boosting with 0.66 and 0.64 scores, respectively. Furthermore, the conducted comparative studies confirm the effectiveness of the proposed approach with improvement of 9% and 12% over BoVWs and AlexNet based features, respectively. Crown Copyright © 2018. Published by Elsevier Ltd. All rights reserved.

Representation learning: a unified deep learning framework for automatic prostate MR segmentation.

PubMed

Liao, Shu; Gao, Yaozong; Oto, Aytekin; Shen, Dinggang

2013-01-01

Image representation plays an important role in medical image analysis. The key to the success of different medical image analysis algorithms is heavily dependent on how we represent the input data, namely features used to characterize the input image. In the literature, feature engineering remains as an active research topic, and many novel hand-crafted features are designed such as Haar wavelet, histogram of oriented gradient, and local binary patterns. However, such features are not designed with the guidance of the underlying dataset at hand. To this end, we argue that the most effective features should be designed in a learning based manner, namely representation learning, which can be adapted to different patient datasets at hand. In this paper, we introduce a deep learning framework to achieve this goal. Specifically, a stacked independent subspace analysis (ISA) network is adopted to learn the most effective features in a hierarchical and unsupervised manner. The learnt features are adapted to the dataset at hand and encode high level semantic anatomical information. The proposed method is evaluated on the application of automatic prostate MR segmentation. Experimental results show that significant segmentation accuracy improvement can be achieved by the proposed deep learning method compared to other state-of-the-art segmentation approaches.

Assessing the Linguistic Productivity of Unsupervised Deep Neural Networks

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

Phillips, Lawrence A.; Hodas, Nathan O.

Increasingly, cognitive scientists have demonstrated interest in applying tools from deep learning. One use for deep learning is in language acquisition where it is useful to know if a linguistic phenomenon can be learned through domain-general means. To assess whether unsupervised deep learning is appropriate, we first pose a smaller question: Can unsupervised neural networks apply linguistic rules productively, using them in novel situations. We draw from the literature on determiner/noun productivity by training an unsupervised, autoencoder network measuring its ability to combine nouns with determiners. Our simple autoencoder creates combinations it has not previously encountered, displaying a degree ofmore » overlap similar to actual children. While this preliminary work does not provide conclusive evidence for productivity, it warrants further investigation with more complex models. Further, this work helps lay the foundations for future collaboration between the deep learning and cognitive science communities.« less

A Physics-Based Deep Learning Approach to Shadow Invariant Representations of Hyperspectral Images.

PubMed

Windrim, Lloyd; Ramakrishnan, Rishi; Melkumyan, Arman; Murphy, Richard J

2018-02-01

This paper proposes the Relit Spectral Angle-Stacked Autoencoder, a novel unsupervised feature learning approach for mapping pixel reflectances to illumination invariant encodings. This work extends the Spectral Angle-Stacked Autoencoder so that it can learn a shadow-invariant mapping. The method is inspired by a deep learning technique, Denoising Autoencoders, with the incorporation of a physics-based model for illumination such that the algorithm learns a shadow invariant mapping without the need for any labelled training data, additional sensors, a priori knowledge of the scene or the assumption of Planckian illumination. The method is evaluated using datasets captured from several different cameras, with experiments to demonstrate the illumination invariance of the features and how they can be used practically to improve the performance of high-level perception algorithms that operate on images acquired outdoors.

Unsupervised spatiotemporal analysis of fMRI data using graph-based visualizations of self-organizing maps.

PubMed

Katwal, Santosh B; Gore, John C; Marois, Rene; Rogers, Baxter P

2013-09-01

We present novel graph-based visualizations of self-organizing maps for unsupervised functional magnetic resonance imaging (fMRI) analysis. A self-organizing map is an artificial neural network model that transforms high-dimensional data into a low-dimensional (often a 2-D) map using unsupervised learning. However, a postprocessing scheme is necessary to correctly interpret similarity between neighboring node prototypes (feature vectors) on the output map and delineate clusters and features of interest in the data. In this paper, we used graph-based visualizations to capture fMRI data features based upon 1) the distribution of data across the receptive fields of the prototypes (density-based connectivity); and 2) temporal similarities (correlations) between the prototypes (correlation-based connectivity). We applied this approach to identify task-related brain areas in an fMRI reaction time experiment involving a visuo-manual response task, and we correlated the time-to-peak of the fMRI responses in these areas with reaction time. Visualization of self-organizing maps outperformed independent component analysis and voxelwise univariate linear regression analysis in identifying and classifying relevant brain regions. We conclude that the graph-based visualizations of self-organizing maps help in advanced visualization of cluster boundaries in fMRI data enabling the separation of regions with small differences in the timings of their brain responses.

Unsupervised Learning and Pattern Recognition of Biological Data Structures with Density Functional Theory and Machine Learning.

PubMed

Chen, Chien-Chang; Juan, Hung-Hui; Tsai, Meng-Yuan; Lu, Henry Horng-Shing

2018-01-11

By introducing the methods of machine learning into the density functional theory, we made a detour for the construction of the most probable density function, which can be estimated by learning relevant features from the system of interest. Using the properties of universal functional, the vital core of density functional theory, the most probable cluster numbers and the corresponding cluster boundaries in a studying system can be simultaneously and automatically determined and the plausibility is erected on the Hohenberg-Kohn theorems. For the method validation and pragmatic applications, interdisciplinary problems from physical to biological systems were enumerated. The amalgamation of uncharged atomic clusters validated the unsupervised searching process of the cluster numbers and the corresponding cluster boundaries were exhibited likewise. High accurate clustering results of the Fisher's iris dataset showed the feasibility and the flexibility of the proposed scheme. Brain tumor detections from low-dimensional magnetic resonance imaging datasets and segmentations of high-dimensional neural network imageries in the Brainbow system were also used to inspect the method practicality. The experimental results exhibit the successful connection between the physical theory and the machine learning methods and will benefit the clinical diagnoses.

Convolutional neural network features based change detection in satellite images

NASA Astrophysics Data System (ADS)

Mohammed El Amin, Arabi; Liu, Qingjie; Wang, Yunhong

2016-07-01

With the popular use of high resolution remote sensing (HRRS) satellite images, a huge research efforts have been placed on change detection (CD) problem. An effective feature selection method can significantly boost the final result. While hand-designed features have proven difficulties to design features that effectively capture high and mid-level representations, the recent developments in machine learning (Deep Learning) omit this problem by learning hierarchical representation in an unsupervised manner directly from data without human intervention. In this letter, we propose approaching the change detection problem from a feature learning perspective. A novel deep Convolutional Neural Networks (CNN) features based HR satellite images change detection method is proposed. The main guideline is to produce a change detection map directly from two images using a pretrained CNN. This method can omit the limited performance of hand-crafted features. Firstly, CNN features are extracted through different convolutional layers. Then, a concatenation step is evaluated after an normalization step, resulting in a unique higher dimensional feature map. Finally, a change map was computed using pixel-wise Euclidean distance. Our method has been validated on real bitemporal HRRS satellite images according to qualitative and quantitative analyses. The results obtained confirm the interest of the proposed method.

Learning spatially coherent properties of the visual world in connectionist networks

NASA Astrophysics Data System (ADS)

Becker, Suzanna; Hinton, Geoffrey E.

1991-10-01

In the unsupervised learning paradigm, a network of neuron-like units is presented with an ensemble of input patterns from a structured environment, such as the visual world, and learns to represent the regularities in that input. The major goal in developing unsupervised learning algorithms is to find objective functions that characterize the quality of the network's representation without explicitly specifying the desired outputs of any of the units. The sort of objective functions considered cause a unit to become tuned to spatially coherent features of visual images (such as texture, depth, shading, and surface orientation), by learning to predict the outputs of other units which have spatially adjacent receptive fields. Simulations show that using an information-theoretic algorithm called IMAX, a network can be trained to represent depth by observing random dot stereograms of surfaces with continuously varying disparities. Once a layer of depth-tuned units has developed, subsequent layers are trained to perform surface interpolation of curved surfaces, by learning to predict the depth of one image region based on depth measurements in surrounding regions. An extension of the basic model allows a population of competing neurons to learn a distributed code for disparity, which naturally gives rise to a representation of discontinuities.

Learned filters for object detection in multi-object visual tracking

NASA Astrophysics Data System (ADS)

Stamatescu, Victor; Wong, Sebastien; McDonnell, Mark D.; Kearney, David

2016-05-01

We investigate the application of learned convolutional filters in multi-object visual tracking. The filters were learned in both a supervised and unsupervised manner from image data using artificial neural networks. This work follows recent results in the field of machine learning that demonstrate the use learned filters for enhanced object detection and classification. Here we employ a track-before-detect approach to multi-object tracking, where tracking guides the detection process. The object detection provides a probabilistic input image calculated by selecting from features obtained using banks of generative or discriminative learned filters. We present a systematic evaluation of these convolutional filters using a real-world data set that examines their performance as generic object detectors.

Context-Aware Local Binary Feature Learning for Face Recognition.

PubMed

Duan, Yueqi; Lu, Jiwen; Feng, Jianjiang; Zhou, Jie

2018-05-01

In this paper, we propose a context-aware local binary feature learning (CA-LBFL) method for face recognition. Unlike existing learning-based local face descriptors such as discriminant face descriptor (DFD) and compact binary face descriptor (CBFD) which learn each feature code individually, our CA-LBFL exploits the contextual information of adjacent bits by constraining the number of shifts from different binary bits, so that more robust information can be exploited for face representation. Given a face image, we first extract pixel difference vectors (PDV) in local patches, and learn a discriminative mapping in an unsupervised manner to project each pixel difference vector into a context-aware binary vector. Then, we perform clustering on the learned binary codes to construct a codebook, and extract a histogram feature for each face image with the learned codebook as the final representation. In order to exploit local information from different scales, we propose a context-aware local binary multi-scale feature learning (CA-LBMFL) method to jointly learn multiple projection matrices for face representation. To make the proposed methods applicable for heterogeneous face recognition, we present a coupled CA-LBFL (C-CA-LBFL) method and a coupled CA-LBMFL (C-CA-LBMFL) method to reduce the modality gap of corresponding heterogeneous faces in the feature level, respectively. Extensive experimental results on four widely used face datasets clearly show that our methods outperform most state-of-the-art face descriptors.

Slow feature analysis: unsupervised learning of invariances.

PubMed

Wiskott, Laurenz; Sejnowski, Terrence J

2002-04-01

Invariant features of temporally varying signals are useful for analysis and classification. Slow feature analysis (SFA) is a new method for learning invariant or slowly varying features from a vectorial input signal. It is based on a nonlinear expansion of the input signal and application of principal component analysis to this expanded signal and its time derivative. It is guaranteed to find the optimal solution within a family of functions directly and can learn to extract a large number of decorrelated features, which are ordered by their degree of invariance. SFA can be applied hierarchically to process high-dimensional input signals and extract complex features. SFA is applied first to complex cell tuning properties based on simple cell output, including disparity and motion. Then more complicated input-output functions are learned by repeated application of SFA. Finally, a hierarchical network of SFA modules is presented as a simple model of the visual system. The same unstructured network can learn translation, size, rotation, contrast, or, to a lesser degree, illumination invariance for one-dimensional objects, depending on only the training stimulus. Surprisingly, only a few training objects suffice to achieve good generalization to new objects. The generated representation is suitable for object recognition. Performance degrades if the network is trained to learn multiple invariances simultaneously.

Teacher and learner: Supervised and unsupervised learning in communities.

PubMed

Shafto, Michael G; Seifert, Colleen M

2015-01-01

How far can teaching methods go to enhance learning? Optimal methods of teaching have been considered in research on supervised and unsupervised learning. Locally optimal methods are usually hybrids of teaching and self-directed approaches. The costs and benefits of specific methods have been shown to depend on the structure of the learning task, the learners, the teachers, and the environment.

Exploiting Redundancy for Flexible Behavior: Unsupervised Learning in a Modular Sensorimotor Control Architecture

ERIC Educational Resources Information Center

Butz, Martin V.; Herbort, Oliver; Hoffmann, Joachim

2007-01-01

Autonomously developing organisms face several challenges when learning reaching movements. First, motor control is learned unsupervised or self-supervised. Second, knowledge of sensorimotor contingencies is acquired in contexts in which action consequences unfold in time. Third, motor redundancies must be resolved. To solve all 3 of these…

A Novel Unsupervised Adaptive Learning Method for Long-Term Electromyography (EMG) Pattern Recognition

PubMed Central

Huang, Qi; Yang, Dapeng; Jiang, Li; Zhang, Huajie; Liu, Hong; Kotani, Kiyoshi

2017-01-01

Performance degradation will be caused by a variety of interfering factors for pattern recognition-based myoelectric control methods in the long term. This paper proposes an adaptive learning method with low computational cost to mitigate the effect in unsupervised adaptive learning scenarios. We presents a particle adaptive classifier (PAC), by constructing a particle adaptive learning strategy and universal incremental least square support vector classifier (LS-SVC). We compared PAC performance with incremental support vector classifier (ISVC) and non-adapting SVC (NSVC) in a long-term pattern recognition task in both unsupervised and supervised adaptive learning scenarios. Retraining time cost and recognition accuracy were compared by validating the classification performance on both simulated and realistic long-term EMG data. The classification results of realistic long-term EMG data showed that the PAC significantly decreased the performance degradation in unsupervised adaptive learning scenarios compared with NSVC (9.03% ± 2.23%, p < 0.05) and ISVC (13.38% ± 2.62%, p = 0.001), and reduced the retraining time cost compared with ISVC (2 ms per updating cycle vs. 50 ms per updating cycle). PMID:28608824

A Novel Unsupervised Adaptive Learning Method for Long-Term Electromyography (EMG) Pattern Recognition.

PubMed

Huang, Qi; Yang, Dapeng; Jiang, Li; Zhang, Huajie; Liu, Hong; Kotani, Kiyoshi

2017-06-13

Performance degradation will be caused by a variety of interfering factors for pattern recognition-based myoelectric control methods in the long term. This paper proposes an adaptive learning method with low computational cost to mitigate the effect in unsupervised adaptive learning scenarios. We presents a particle adaptive classifier (PAC), by constructing a particle adaptive learning strategy and universal incremental least square support vector classifier (LS-SVC). We compared PAC performance with incremental support vector classifier (ISVC) and non-adapting SVC (NSVC) in a long-term pattern recognition task in both unsupervised and supervised adaptive learning scenarios. Retraining time cost and recognition accuracy were compared by validating the classification performance on both simulated and realistic long-term EMG data. The classification results of realistic long-term EMG data showed that the PAC significantly decreased the performance degradation in unsupervised adaptive learning scenarios compared with NSVC (9.03% ± 2.23%, p < 0.05) and ISVC (13.38% ± 2.62%, p = 0.001), and reduced the retraining time cost compared with ISVC (2 ms per updating cycle vs. 50 ms per updating cycle).

Characterization of electroencephalography signals for estimating saliency features in videos.

PubMed

Liang, Zhen; Hamada, Yasuyuki; Oba, Shigeyuki; Ishii, Shin

2018-05-12

Understanding the functions of the visual system has been one of the major targets in neuroscience formany years. However, the relation between spontaneous brain activities and visual saliency in natural stimuli has yet to be elucidated. In this study, we developed an optimized machine learning-based decoding model to explore the possible relationships between the electroencephalography (EEG) characteristics and visual saliency. The optimal features were extracted from the EEG signals and saliency map which was computed according to an unsupervised saliency model ( Tavakoli and Laaksonen, 2017). Subsequently, various unsupervised feature selection/extraction techniques were examined using different supervised regression models. The robustness of the presented model was fully verified by means of ten-fold or nested cross validation procedure, and promising results were achieved in the reconstruction of saliency features based on the selected EEG characteristics. Through the successful demonstration of using EEG characteristics to predict the real-time saliency distribution in natural videos, we suggest the feasibility of quantifying visual content through measuring brain activities (EEG signals) in real environments, which would facilitate the understanding of cortical involvement in the processing of natural visual stimuli and application developments motivated by human visual processing. Copyright © 2018 Elsevier Ltd. All rights reserved.

Unsupervised Learning (Clustering) of Odontocete Echolocation Clicks

DTIC Science & Technology

2015-09-30

of their bandwidth. Results on Risso’s dolphins (Grampus griseus), Pacific white-sided dolphins (Lagenorhynchus obliquidens), and Cuvier’s beaked...acoustic encounters to see which ones appeared to be closely related to one another. We noted that some of the Pacific white-sided and Risso’s dolphin ...should be clusterable. The group of odontocetes that we cannot label reliably by their acoustic features, primarily common dolphins (Delphinus spp

Low-dimensional dynamical characterization of human performance of cancer patients using motion data.

PubMed

Hasnain, Zaki; Li, Ming; Dorff, Tanya; Quinn, David; Ueno, Naoto T; Yennu, Sriram; Kolatkar, Anand; Shahabi, Cyrus; Nocera, Luciano; Nieva, Jorge; Kuhn, Peter; Newton, Paul K

2018-05-18

Biomechanical characterization of human performance with respect to fatigue and fitness is relevant in many settings, however is usually limited to either fully qualitative assessments or invasive methods which require a significant experimental setup consisting of numerous sensors, force plates, and motion detectors. Qualitative assessments are difficult to standardize due to their intrinsic subjective nature, on the other hand, invasive methods provide reliable metrics but are not feasible for large scale applications. Presented here is a dynamical toolset for detecting performance groups using a non-invasive system based on the Microsoft Kinect motion capture sensor, and a case study of 37 cancer patients performing two clinically monitored tasks before and after therapy regimens. Dynamical features are extracted from the motion time series data and evaluated based on their ability to i) cluster patients into coherent fitness groups using unsupervised learning algorithms and to ii) predict Eastern Cooperative Oncology Group performance status via supervised learning. The unsupervised patient clustering is comparable to clustering based on physician assigned Eastern Cooperative Oncology Group status in that they both have similar concordance with change in weight before and after therapy as well as unexpected hospitalizations throughout the study. The extracted dynamical features can predict physician, coordinator, and patient Eastern Cooperative Oncology Group status with an accuracy of approximately 80%. The non-invasive Microsoft Kinect sensor and the proposed dynamical toolset comprised of data preprocessing, feature extraction, dimensionality reduction, and machine learning offers a low-cost and general method for performance segregation and can complement existing qualitative clinical assessments. Copyright © 2018 Elsevier Ltd. All rights reserved.

Autonomous learning in gesture recognition by using lobe component analysis

NASA Astrophysics Data System (ADS)

Lu, Jian; Weng, Juyang

2007-02-01

Gesture recognition is a new human-machine interface method implemented by pattern recognition(PR).In order to assure robot safety when gesture is used in robot control, it is required to implement the interface reliably and accurately. Similar with other PR applications, 1) feature selection (or model establishment) and 2) training from samples, affect the performance of gesture recognition largely. For 1), a simple model with 6 feature points at shoulders, elbows, and hands, is established. The gestures to be recognized are restricted to still arm gestures, and the movement of arms is not considered. These restrictions are to reduce the misrecognition, but are not so unreasonable. For 2), a new biological network method, called lobe component analysis(LCA), is used in unsupervised learning. Lobe components, corresponding to high-concentrations in probability of the neuronal input, are orientation selective cells follow Hebbian rule and lateral inhibition. Due to the advantage of LCA method for balanced learning between global and local features, large amount of samples can be used in learning efficiently.

Exploiting Secondary Sources for Unsupervised Record Linkage

DTIC Science & Technology

2004-01-01

paper, we present an extension to Apollo’s active learning component to Report Documentation Page Form ApprovedOMB No. 0704-0188 Public reporting...Sources address the issue of user involvement. Using secondary sources, a system can autonomously answer questions posed by its active learning component...over, we present how Apollo utilizes the identified sec- ondary sources in an unsupervised active learning pro- cess. Apollo’s learning algorithm

Modelling unsupervised online-learning of artificial grammars: linking implicit and statistical learning.

PubMed

Rohrmeier, Martin A; Cross, Ian

2014-07-01

Humans rapidly learn complex structures in various domains. Findings of above-chance performance of some untrained control groups in artificial grammar learning studies raise questions about the extent to which learning can occur in an untrained, unsupervised testing situation with both correct and incorrect structures. The plausibility of unsupervised online-learning effects was modelled with n-gram, chunking and simple recurrent network models. A novel evaluation framework was applied, which alternates forced binary grammaticality judgments and subsequent learning of the same stimulus. Our results indicate a strong online learning effect for n-gram and chunking models and a weaker effect for simple recurrent network models. Such findings suggest that online learning is a plausible effect of statistical chunk learning that is possible when ungrammatical sequences contain a large proportion of grammatical chunks. Such common effects of continuous statistical learning may underlie statistical and implicit learning paradigms and raise implications for study design and testing methodologies. Copyright © 2014 Elsevier Inc. All rights reserved.

Image segmentation using fuzzy LVQ clustering networks

NASA Technical Reports Server (NTRS)

Tsao, Eric Chen-Kuo; Bezdek, James C.; Pal, Nikhil R.

1992-01-01

In this note we formulate image segmentation as a clustering problem. Feature vectors extracted from a raw image are clustered into subregions, thereby segmenting the image. A fuzzy generalization of a Kohonen learning vector quantization (LVQ) which integrates the Fuzzy c-Means (FCM) model with the learning rate and updating strategies of the LVQ is used for this task. This network, which segments images in an unsupervised manner, is thus related to the FCM optimization problem. Numerical examples on photographic and magnetic resonance images are given to illustrate this approach to image segmentation.

Unsupervised algorithms for intrusion detection and identification in wireless ad hoc sensor networks

NASA Astrophysics Data System (ADS)

Hortos, William S.

2009-05-01

In previous work by the author, parameters across network protocol layers were selected as features in supervised algorithms that detect and identify certain intrusion attacks on wireless ad hoc sensor networks (WSNs) carrying multisensor data. The algorithms improved the residual performance of the intrusion prevention measures provided by any dynamic key-management schemes and trust models implemented among network nodes. The approach of this paper does not train algorithms on the signature of known attack traffic, but, instead, the approach is based on unsupervised anomaly detection techniques that learn the signature of normal network traffic. Unsupervised learning does not require the data to be labeled or to be purely of one type, i.e., normal or attack traffic. The approach can be augmented to add any security attributes and quantified trust levels, established during data exchanges among nodes, to the set of cross-layer features from the WSN protocols. A two-stage framework is introduced for the security algorithms to overcome the problems of input size and resource constraints. The first stage is an unsupervised clustering algorithm which reduces the payload of network data packets to a tractable size. The second stage is a traditional anomaly detection algorithm based on a variation of support vector machines (SVMs), whose efficiency is improved by the availability of data in the packet payload. In the first stage, selected algorithms are adapted to WSN platforms to meet system requirements for simple parallel distributed computation, distributed storage and data robustness. A set of mobile software agents, acting like an ant colony in securing the WSN, are distributed at the nodes to implement the algorithms. The agents move among the layers involved in the network response to the intrusions at each active node and trustworthy neighborhood, collecting parametric values and executing assigned decision tasks. This minimizes the need to move large amounts of audit-log data through resource-limited nodes and locates routines closer to that data. Performance of the unsupervised algorithms is evaluated against the network intrusions of black hole, flooding, Sybil and other denial-of-service attacks in simulations of published scenarios. Results for scenarios with intentionally malfunctioning sensors show the robustness of the two-stage approach to intrusion anomalies.

Combining Unsupervised and Supervised Classification to Build User Models for Exploratory Learning Environments

ERIC Educational Resources Information Center

Amershi, Saleema; Conati, Cristina

2009-01-01

In this paper, we present a data-based user modeling framework that uses both unsupervised and supervised classification to build student models for exploratory learning environments. We apply the framework to build student models for two different learning environments and using two different data sources (logged interface and eye-tracking data).…

Unsupervised Transfer Learning via Multi-Scale Convolutional Sparse Coding for Biomedical Applications

PubMed Central

Chang, Hang; Han, Ju; Zhong, Cheng; Snijders, Antoine M.; Mao, Jian-Hua

2017-01-01

The capabilities of (I) learning transferable knowledge across domains; and (II) fine-tuning the pre-learned base knowledge towards tasks with considerably smaller data scale are extremely important. Many of the existing transfer learning techniques are supervised approaches, among which deep learning has the demonstrated power of learning domain transferrable knowledge with large scale network trained on massive amounts of labeled data. However, in many biomedical tasks, both the data and the corresponding label can be very limited, where the unsupervised transfer learning capability is urgently needed. In this paper, we proposed a novel multi-scale convolutional sparse coding (MSCSC) method, that (I) automatically learns filter banks at different scales in a joint fashion with enforced scale-specificity of learned patterns; and (II) provides an unsupervised solution for learning transferable base knowledge and fine-tuning it towards target tasks. Extensive experimental evaluation of MSCSC demonstrates the effectiveness of the proposed MSCSC in both regular and transfer learning tasks in various biomedical domains. PMID:28129148

Surface mapping via unsupervised classification of remote sensing: application to MESSENGER/MASCS and DAWN/VIRS data.

NASA Astrophysics Data System (ADS)

D'Amore, M.; Le Scaon, R.; Helbert, J.; Maturilli, A.

2017-12-01

Machine-learning achieved unprecedented results in high-dimensional data processing tasks with wide applications in various fields. Due to the growing number of complex nonlinear systems that have to be investigated in science and the bare raw size of data nowadays available, ML offers the unique ability to extract knowledge, regardless the specific application field. Examples are image segmentation, supervised/unsupervised/ semi-supervised classification, feature extraction, data dimensionality analysis/reduction.The MASCS instrument has mapped Mercury surface in the 400-1145 nm wavelength range during orbital observations by the MESSENGER spacecraft. We have conducted k-means unsupervised hierarchical clustering to identify and characterize spectral units from MASCS observations. The results display a dichotomy: a polar and equatorial units, possibly linked to compositional differences or weathering due to irradiation. To explore possible relations between composition and spectral behavior, we have compared the spectral provinces with elemental abundance maps derived from MESSENGER's X-Ray Spectrometer (XRS).For the Vesta application on DAWN Visible and infrared spectrometer (VIR) data, we explored several Machine Learning techniques: image segmentation method, stream algorithm and hierarchical clustering.The algorithm successfully separates the Olivine outcrops around two craters on Vesta's surface [1]. New maps summarizing the spectral and chemical signature of the surface could be automatically produced.We conclude that instead of hand digging in data, scientist could choose a subset of algorithms with well known feature (i.e. efficacy on the particular problem, speed, accuracy) and focus their effort in understanding what important characteristic of the groups found in the data mean. [1] E Ammannito et al. "Olivine in an unexpected location on Vesta's surface". In: Nature 504.7478 (2013), pp. 122-125.

Accuracy of latent-variable estimation in Bayesian semi-supervised learning.

PubMed

Yamazaki, Keisuke

2015-09-01

Hierarchical probabilistic models, such as Gaussian mixture models, are widely used for unsupervised learning tasks. These models consist of observable and latent variables, which represent the observable data and the underlying data-generation process, respectively. Unsupervised learning tasks, such as cluster analysis, are regarded as estimations of latent variables based on the observable ones. The estimation of latent variables in semi-supervised learning, where some labels are observed, will be more precise than that in unsupervised, and one of the concerns is to clarify the effect of the labeled data. However, there has not been sufficient theoretical analysis of the accuracy of the estimation of latent variables. In a previous study, a distribution-based error function was formulated, and its asymptotic form was calculated for unsupervised learning with generative models. It has been shown that, for the estimation of latent variables, the Bayes method is more accurate than the maximum-likelihood method. The present paper reveals the asymptotic forms of the error function in Bayesian semi-supervised learning for both discriminative and generative models. The results show that the generative model, which uses all of the given data, performs better when the model is well specified. Copyright © 2015 Elsevier Ltd. All rights reserved.

Unsupervised learning in probabilistic neural networks with multi-state metal-oxide memristive synapses

NASA Astrophysics Data System (ADS)

Serb, Alexander; Bill, Johannes; Khiat, Ali; Berdan, Radu; Legenstein, Robert; Prodromakis, Themis

2016-09-01

In an increasingly data-rich world the need for developing computing systems that cannot only process, but ideally also interpret big data is becoming continuously more pressing. Brain-inspired concepts have shown great promise towards addressing this need. Here we demonstrate unsupervised learning in a probabilistic neural network that utilizes metal-oxide memristive devices as multi-state synapses. Our approach can be exploited for processing unlabelled data and can adapt to time-varying clusters that underlie incoming data by supporting the capability of reversible unsupervised learning. The potential of this work is showcased through the demonstration of successful learning in the presence of corrupted input data and probabilistic neurons, thus paving the way towards robust big-data processors.

Using Wikipedia to learn semantic feature representations of concrete concepts in neuroimaging experiments

PubMed Central

Pereira, Francisco; Botvinick, Matthew; Detre, Greg

2012-01-01

In this paper we show that a corpus of a few thousand Wikipedia articles about concrete or visualizable concepts can be used to produce a low-dimensional semantic feature representation of those concepts. The purpose of such a representation is to serve as a model of the mental context of a subject during functional magnetic resonance imaging (fMRI) experiments. A recent study [19] showed that it was possible to predict fMRI data acquired while subjects thought about a concrete concept, given a representation of those concepts in terms of semantic features obtained with human supervision. We use topic models on our corpus to learn semantic features from text in an unsupervised manner, and show that those features can outperform those in [19] in demanding 12-way and 60-way classification tasks. We also show that these features can be used to uncover similarity relations in brain activation for different concepts which parallel those relations in behavioral data from human subjects. PMID:23243317

Towards a new classification of stable phase schizophrenia into major and simple neuro-cognitive psychosis: Results of unsupervised machine learning analysis.

PubMed

Kanchanatawan, Buranee; Sriswasdi, Sira; Thika, Supaksorn; Stoyanov, Drozdstoy; Sirivichayakul, Sunee; Carvalho, André F; Geffard, Michel; Maes, Michael

2018-05-23

Deficit schizophrenia, as defined by the Schedule for Deficit Syndrome, may represent a distinct diagnostic class defined by neurocognitive impairments coupled with changes in IgA/IgM responses to tryptophan catabolites (TRYCATs). Adequate classifications should be based on supervised and unsupervised learning rather than on consensus criteria. This study used machine learning as means to provide a more accurate classification of patients with stable phase schizophrenia. We found that using negative symptoms as discriminatory variables, schizophrenia patients may be divided into two distinct classes modelled by (A) impairments in IgA/IgM responses to noxious and generally more protective tryptophan catabolites, (B) impairments in episodic and semantic memory, paired associative learning and false memory creation, and (C) psychotic, excitation, hostility, mannerism, negative, and affective symptoms. The first cluster shows increased negative, psychotic, excitation, hostility, mannerism, depression and anxiety symptoms, and more neuroimmune and cognitive disorders and is therefore called "major neurocognitive psychosis" (MNP). The second cluster, called "simple neurocognitive psychosis" (SNP) is discriminated from normal controls by the same features although the impairments are less well developed than in MNP. The latter is additionally externally validated by lowered quality of life, body mass (reflecting a leptosome body type), and education (reflecting lower cognitive reserve). Previous distinctions including "type 1" (positive)/"type 2" (negative) and DSM-IV-TR (eg, paranoid) schizophrenia could not be validated using machine learning techniques. Previous names of the illness, including schizophrenia, are not very adequate because they do not describe the features of the illness, namely, interrelated neuroimmune, cognitive, and clinical features. Stable-phase schizophrenia consists of 2 relevant qualitatively distinct categories or nosological entities with SNP being a less well-developed phenotype, while MNP is the full blown phenotype or core illness. Major neurocognitive psychosis and SNP should be added to the DSM-5 and incorporated into the Research Domain Criteria project. © 2018 John Wiley & Sons, Ltd.

Large-scale Exploration of Neuronal Morphologies Using Deep Learning and Augmented Reality.

PubMed

Li, Zhongyu; Butler, Erik; Li, Kang; Lu, Aidong; Ji, Shuiwang; Zhang, Shaoting

2018-02-12

Recently released large-scale neuron morphological data has greatly facilitated the research in neuroinformatics. However, the sheer volume and complexity of these data pose significant challenges for efficient and accurate neuron exploration. In this paper, we propose an effective retrieval framework to address these problems, based on frontier techniques of deep learning and binary coding. For the first time, we develop a deep learning based feature representation method for the neuron morphological data, where the 3D neurons are first projected into binary images and then learned features using an unsupervised deep neural network, i.e., stacked convolutional autoencoders (SCAEs). The deep features are subsequently fused with the hand-crafted features for more accurate representation. Considering the exhaustive search is usually very time-consuming in large-scale databases, we employ a novel binary coding method to compress feature vectors into short binary codes. Our framework is validated on a public data set including 58,000 neurons, showing promising retrieval precision and efficiency compared with state-of-the-art methods. In addition, we develop a novel neuron visualization program based on the techniques of augmented reality (AR), which can help users take a deep exploration of neuron morphologies in an interactive and immersive manner.

Feature learning and change feature classification based on deep learning for ternary change detection in SAR images

NASA Astrophysics Data System (ADS)

Gong, Maoguo; Yang, Hailun; Zhang, Puzhao

2017-07-01

Ternary change detection aims to detect changes and group the changes into positive change and negative change. It is of great significance in the joint interpretation of spatial-temporal synthetic aperture radar images. In this study, sparse autoencoder, convolutional neural networks (CNN) and unsupervised clustering are combined to solve ternary change detection problem without any supervison. Firstly, sparse autoencoder is used to transform log-ratio difference image into a suitable feature space for extracting key changes and suppressing outliers and noise. And then the learned features are clustered into three classes, which are taken as the pseudo labels for training a CNN model as change feature classifier. The reliable training samples for CNN are selected from the feature maps learned by sparse autoencoder with certain selection rules. Having training samples and the corresponding pseudo labels, the CNN model can be trained by using back propagation with stochastic gradient descent. During its training procedure, CNN is driven to learn the concept of change, and more powerful model is established to distinguish different types of changes. Unlike the traditional methods, the proposed framework integrates the merits of sparse autoencoder and CNN to learn more robust difference representations and the concept of change for ternary change detection. Experimental results on real datasets validate the effectiveness and superiority of the proposed framework.

Exploring supervised and unsupervised methods to detect topics in biomedical text

PubMed Central

Lee, Minsuk; Wang, Weiqing; Yu, Hong

2006-01-01

Background Topic detection is a task that automatically identifies topics (e.g., "biochemistry" and "protein structure") in scientific articles based on information content. Topic detection will benefit many other natural language processing tasks including information retrieval, text summarization and question answering; and is a necessary step towards the building of an information system that provides an efficient way for biologists to seek information from an ocean of literature. Results We have explored the methods of Topic Spotting, a task of text categorization that applies the supervised machine-learning technique naïve Bayes to assign automatically a document into one or more predefined topics; and Topic Clustering, which apply unsupervised hierarchical clustering algorithms to aggregate documents into clusters such that each cluster represents a topic. We have applied our methods to detect topics of more than fifteen thousand of articles that represent over sixteen thousand entries in the Online Mendelian Inheritance in Man (OMIM) database. We have explored bag of words as the features. Additionally, we have explored semantic features; namely, the Medical Subject Headings (MeSH) that are assigned to the MEDLINE records, and the Unified Medical Language System (UMLS) semantic types that correspond to the MeSH terms, in addition to bag of words, to facilitate the tasks of topic detection. Our results indicate that incorporating the MeSH terms and the UMLS semantic types as additional features enhances the performance of topic detection and the naïve Bayes has the highest accuracy, 66.4%, for predicting the topic of an OMIM article as one of the total twenty-five topics. Conclusion Our results indicate that the supervised topic spotting methods outperformed the unsupervised topic clustering; on the other hand, the unsupervised topic clustering methods have the advantages of being robust and applicable in real world settings. PMID:16539745

A Hybrid Supervised/Unsupervised Machine Learning Approach to Solar Flare Prediction

NASA Astrophysics Data System (ADS)

Benvenuto, Federico; Piana, Michele; Campi, Cristina; Massone, Anna Maria

2018-01-01

This paper introduces a novel method for flare forecasting, combining prediction accuracy with the ability to identify the most relevant predictive variables. This result is obtained by means of a two-step approach: first, a supervised regularization method for regression, namely, LASSO is applied, where a sparsity-enhancing penalty term allows the identification of the significance with which each data feature contributes to the prediction; then, an unsupervised fuzzy clustering technique for classification, namely, Fuzzy C-Means, is applied, where the regression outcome is partitioned through the minimization of a cost function and without focusing on the optimization of a specific skill score. This approach is therefore hybrid, since it combines supervised and unsupervised learning; realizes classification in an automatic, skill-score-independent way; and provides effective prediction performances even in the case of imbalanced data sets. Its prediction power is verified against NOAA Space Weather Prediction Center data, using as a test set, data in the range between 1996 August and 2010 December and as training set, data in the range between 1988 December and 1996 June. To validate the method, we computed several skill scores typically utilized in flare prediction and compared the values provided by the hybrid approach with the ones provided by several standard (non-hybrid) machine learning methods. The results showed that the hybrid approach performs classification better than all other supervised methods and with an effectiveness comparable to the one of clustering methods; but, in addition, it provides a reliable ranking of the weights with which the data properties contribute to the forecast.

Learning from label proportions in brain-computer interfaces: Online unsupervised learning with guarantees.

PubMed

Hübner, David; Verhoeven, Thibault; Schmid, Konstantin; Müller, Klaus-Robert; Tangermann, Michael; Kindermans, Pieter-Jan

2017-01-01

Using traditional approaches, a brain-computer interface (BCI) requires the collection of calibration data for new subjects prior to online use. Calibration time can be reduced or eliminated e.g., by subject-to-subject transfer of a pre-trained classifier or unsupervised adaptive classification methods which learn from scratch and adapt over time. While such heuristics work well in practice, none of them can provide theoretical guarantees. Our objective is to modify an event-related potential (ERP) paradigm to work in unison with the machine learning decoder, and thus to achieve a reliable unsupervised calibrationless decoding with a guarantee to recover the true class means. We introduce learning from label proportions (LLP) to the BCI community as a new unsupervised, and easy-to-implement classification approach for ERP-based BCIs. The LLP estimates the mean target and non-target responses based on known proportions of these two classes in different groups of the data. We present a visual ERP speller to meet the requirements of LLP. For evaluation, we ran simulations on artificially created data sets and conducted an online BCI study with 13 subjects performing a copy-spelling task. Theoretical considerations show that LLP is guaranteed to minimize the loss function similar to a corresponding supervised classifier. LLP performed well in simulations and in the online application, where 84.5% of characters were spelled correctly on average without prior calibration. The continuously adapting LLP classifier is the first unsupervised decoder for ERP BCIs guaranteed to find the optimal decoder. This makes it an ideal solution to avoid tedious calibration sessions. Additionally, LLP works on complementary principles compared to existing unsupervised methods, opening the door for their further enhancement when combined with LLP.

Learning from label proportions in brain-computer interfaces: Online unsupervised learning with guarantees

PubMed Central

Verhoeven, Thibault; Schmid, Konstantin; Müller, Klaus-Robert; Tangermann, Michael; Kindermans, Pieter-Jan

2017-01-01

Objective Using traditional approaches, a brain-computer interface (BCI) requires the collection of calibration data for new subjects prior to online use. Calibration time can be reduced or eliminated e.g., by subject-to-subject transfer of a pre-trained classifier or unsupervised adaptive classification methods which learn from scratch and adapt over time. While such heuristics work well in practice, none of them can provide theoretical guarantees. Our objective is to modify an event-related potential (ERP) paradigm to work in unison with the machine learning decoder, and thus to achieve a reliable unsupervised calibrationless decoding with a guarantee to recover the true class means. Method We introduce learning from label proportions (LLP) to the BCI community as a new unsupervised, and easy-to-implement classification approach for ERP-based BCIs. The LLP estimates the mean target and non-target responses based on known proportions of these two classes in different groups of the data. We present a visual ERP speller to meet the requirements of LLP. For evaluation, we ran simulations on artificially created data sets and conducted an online BCI study with 13 subjects performing a copy-spelling task. Results Theoretical considerations show that LLP is guaranteed to minimize the loss function similar to a corresponding supervised classifier. LLP performed well in simulations and in the online application, where 84.5% of characters were spelled correctly on average without prior calibration. Significance The continuously adapting LLP classifier is the first unsupervised decoder for ERP BCIs guaranteed to find the optimal decoder. This makes it an ideal solution to avoid tedious calibration sessions. Additionally, LLP works on complementary principles compared to existing unsupervised methods, opening the door for their further enhancement when combined with LLP. PMID:28407016

Feature Extraction Using an Unsupervised Neural Network

DTIC Science & Technology

1991-05-03

with this neural netowrk is given and its connection to exploratory projection pursuit methods is established. DD I 2 P JA d 73 EDITIONj Of I NOV 6s...IS OBSOLETE $IN 0102- LF- 014- 6601 SECURITY CLASSIFICATION OF THIS PAGE (When Daoes Enlered) Feature Extraction using an Unsupervised Neural Network

Prediction of enhancer-promoter interactions via natural language processing.

PubMed

Zeng, Wanwen; Wu, Mengmeng; Jiang, Rui

2018-05-09

Precise identification of three-dimensional genome organization, especially enhancer-promoter interactions (EPIs), is important to deciphering gene regulation, cell differentiation and disease mechanisms. Currently, it is a challenging task to distinguish true interactions from other nearby non-interacting ones since the power of traditional experimental methods is limited due to low resolution or low throughput. We propose a novel computational framework EP2vec to assay three-dimensional genomic interactions. We first extract sequence embedding features, defined as fixed-length vector representations learned from variable-length sequences using an unsupervised deep learning method in natural language processing. Then, we train a classifier to predict EPIs using the learned representations in supervised way. Experimental results demonstrate that EP2vec obtains F1 scores ranging from 0.841~ 0.933 on different datasets, which outperforms existing methods. We prove the robustness of sequence embedding features by carrying out sensitivity analysis. Besides, we identify motifs that represent cell line-specific information through analysis of the learned sequence embedding features by adopting attention mechanism. Last, we show that even superior performance with F1 scores 0.889~ 0.940 can be achieved by combining sequence embedding features and experimental features. EP2vec sheds light on feature extraction for DNA sequences of arbitrary lengths and provides a powerful approach for EPIs identification.

Similarity preserving low-rank representation for enhanced data representation and effective subspace learning.

PubMed

Zhang, Zhao; Yan, Shuicheng; Zhao, Mingbo

2014-05-01

Latent Low-Rank Representation (LatLRR) delivers robust and promising results for subspace recovery and feature extraction through mining the so-called hidden effects, but the locality of both similar principal and salient features cannot be preserved in the optimizations. To solve this issue for achieving enhanced performance, a boosted version of LatLRR, referred to as Regularized Low-Rank Representation (rLRR), is proposed through explicitly including an appropriate Laplacian regularization that can maximally preserve the similarity among local features. Resembling LatLRR, rLRR decomposes given data matrix from two directions by seeking a pair of low-rank matrices. But the similarities of principal and salient features can be effectively preserved by rLRR. As a result, the correlated features are well grouped and the robustness of representations is also enhanced. Based on the outputted bi-directional low-rank codes by rLRR, an unsupervised subspace learning framework termed Low-rank Similarity Preserving Projections (LSPP) is also derived for feature learning. The supervised extension of LSPP is also discussed for discriminant subspace learning. The validity of rLRR is examined by robust representation and decomposition of real images. Results demonstrated the superiority of our rLRR and LSPP in comparison to other related state-of-the-art algorithms. Copyright © 2014 Elsevier Ltd. All rights reserved.

Synaptic State Matching: A Dynamical Architecture for Predictive Internal Representation and Feature Detection

PubMed Central

Tavazoie, Saeed

2013-01-01

Here we explore the possibility that a core function of sensory cortex is the generation of an internal simulation of sensory environment in real-time. A logical elaboration of this idea leads to a dynamical neural architecture that oscillates between two fundamental network states, one driven by external input, and the other by recurrent synaptic drive in the absence of sensory input. Synaptic strength is modified by a proposed synaptic state matching (SSM) process that ensures equivalence of spike statistics between the two network states. Remarkably, SSM, operating locally at individual synapses, generates accurate and stable network-level predictive internal representations, enabling pattern completion and unsupervised feature detection from noisy sensory input. SSM is a biologically plausible substrate for learning and memory because it brings together sequence learning, feature detection, synaptic homeostasis, and network oscillations under a single unifying computational framework. PMID:23991161

A Granular Self-Organizing Map for Clustering and Gene Selection in Microarray Data.

PubMed

Ray, Shubhra Sankar; Ganivada, Avatharam; Pal, Sankar K

2016-09-01

A new granular self-organizing map (GSOM) is developed by integrating the concept of a fuzzy rough set with the SOM. While training the GSOM, the weights of a winning neuron and the neighborhood neurons are updated through a modified learning procedure. The neighborhood is newly defined using the fuzzy rough sets. The clusters (granules) evolved by the GSOM are presented to a decision table as its decision classes. Based on the decision table, a method of gene selection is developed. The effectiveness of the GSOM is shown in both clustering samples and developing an unsupervised fuzzy rough feature selection (UFRFS) method for gene selection in microarray data. While the superior results of the GSOM, as compared with the related clustering methods, are provided in terms of β -index, DB-index, Dunn-index, and fuzzy rough entropy, the genes selected by the UFRFS are not only better in terms of classification accuracy and a feature evaluation index, but also statistically more significant than the related unsupervised methods. The C-codes of the GSOM and UFRFS are available online at http://avatharamg.webs.com/software-code.

A Novel Unsupervised Segmentation Quality Evaluation Method for Remote Sensing Images

PubMed Central

Tang, Yunwei; Jing, Linhai; Ding, Haifeng

2017-01-01

The segmentation of a high spatial resolution remote sensing image is a critical step in geographic object-based image analysis (GEOBIA). Evaluating the performance of segmentation without ground truth data, i.e., unsupervised evaluation, is important for the comparison of segmentation algorithms and the automatic selection of optimal parameters. This unsupervised strategy currently faces several challenges in practice, such as difficulties in designing effective indicators and limitations of the spectral values in the feature representation. This study proposes a novel unsupervised evaluation method to quantitatively measure the quality of segmentation results to overcome these problems. In this method, multiple spectral and spatial features of images are first extracted simultaneously and then integrated into a feature set to improve the quality of the feature representation of ground objects. The indicators designed for spatial stratified heterogeneity and spatial autocorrelation are included to estimate the properties of the segments in this integrated feature set. These two indicators are then combined into a global assessment metric as the final quality score. The trade-offs of the combined indicators are accounted for using a strategy based on the Mahalanobis distance, which can be exhibited geometrically. The method is tested on two segmentation algorithms and three testing images. The proposed method is compared with two existing unsupervised methods and a supervised method to confirm its capabilities. Through comparison and visual analysis, the results verified the effectiveness of the proposed method and demonstrated the reliability and improvements of this method with respect to other methods. PMID:29064416

Sparse Coding for N-Gram Feature Extraction and Training for File Fragment Classification

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

Wang, Felix; Quach, Tu-Thach; Wheeler, Jason

File fragment classification is an important step in the task of file carving in digital forensics. In file carving, files must be reconstructed based on their content as a result of their fragmented storage on disk or in memory. Existing methods for classification of file fragments typically use hand-engineered features such as byte histograms or entropy measures. In this paper, we propose an approach using sparse coding that enables automated feature extraction. Sparse coding, or sparse dictionary learning, is an unsupervised learning algorithm, and is capable of extracting features based simply on how well those features can be used tomore » reconstruct the original data. With respect to file fragments, we learn sparse dictionaries for n-grams, continuous sequences of bytes, of different sizes. These dictionaries may then be used to estimate n-gram frequencies for a given file fragment, but for significantly larger n-gram sizes than are typically found in existing methods which suffer from combinatorial explosion. To demonstrate the capability of our sparse coding approach, we used the resulting features to train standard classifiers such as support vector machines (SVMs) over multiple file types. Experimentally, we achieved significantly better classification results with respect to existing methods, especially when the features were used in supplement to existing hand-engineered features.« less

Sparse Coding for N-Gram Feature Extraction and Training for File Fragment Classification

DOE PAGES

Wang, Felix; Quach, Tu-Thach; Wheeler, Jason; ...

2018-04-05

File fragment classification is an important step in the task of file carving in digital forensics. In file carving, files must be reconstructed based on their content as a result of their fragmented storage on disk or in memory. Existing methods for classification of file fragments typically use hand-engineered features such as byte histograms or entropy measures. In this paper, we propose an approach using sparse coding that enables automated feature extraction. Sparse coding, or sparse dictionary learning, is an unsupervised learning algorithm, and is capable of extracting features based simply on how well those features can be used tomore » reconstruct the original data. With respect to file fragments, we learn sparse dictionaries for n-grams, continuous sequences of bytes, of different sizes. These dictionaries may then be used to estimate n-gram frequencies for a given file fragment, but for significantly larger n-gram sizes than are typically found in existing methods which suffer from combinatorial explosion. To demonstrate the capability of our sparse coding approach, we used the resulting features to train standard classifiers such as support vector machines (SVMs) over multiple file types. Experimentally, we achieved significantly better classification results with respect to existing methods, especially when the features were used in supplement to existing hand-engineered features.« less

STDP-based spiking deep convolutional neural networks for object recognition.

PubMed

Kheradpisheh, Saeed Reza; Ganjtabesh, Mohammad; Thorpe, Simon J; Masquelier, Timothée

2018-03-01

Previous studies have shown that spike-timing-dependent plasticity (STDP) can be used in spiking neural networks (SNN) to extract visual features of low or intermediate complexity in an unsupervised manner. These studies, however, used relatively shallow architectures, and only one layer was trainable. Another line of research has demonstrated - using rate-based neural networks trained with back-propagation - that having many layers increases the recognition robustness, an approach known as deep learning. We thus designed a deep SNN, comprising several convolutional (trainable with STDP) and pooling layers. We used a temporal coding scheme where the most strongly activated neurons fire first, and less activated neurons fire later or not at all. The network was exposed to natural images. Thanks to STDP, neurons progressively learned features corresponding to prototypical patterns that were both salient and frequent. Only a few tens of examples per category were required and no label was needed. After learning, the complexity of the extracted features increased along the hierarchy, from edge detectors in the first layer to object prototypes in the last layer. Coding was very sparse, with only a few thousands spikes per image, and in some cases the object category could be reasonably well inferred from the activity of a single higher-order neuron. More generally, the activity of a few hundreds of such neurons contained robust category information, as demonstrated using a classifier on Caltech 101, ETH-80, and MNIST databases. We also demonstrate the superiority of STDP over other unsupervised techniques such as random crops (HMAX) or auto-encoders. Taken together, our results suggest that the combination of STDP with latency coding may be a key to understanding the way that the primate visual system learns, its remarkable processing speed and its low energy consumption. These mechanisms are also interesting for artificial vision systems, particularly for hardware solutions. Copyright © 2017 Elsevier Ltd. All rights reserved.

Enhancing interpretability of automatically extracted machine learning features: application to a RBM-Random Forest system on brain lesion segmentation.

PubMed

Pereira, Sérgio; Meier, Raphael; McKinley, Richard; Wiest, Roland; Alves, Victor; Silva, Carlos A; Reyes, Mauricio

2018-02-01

Machine learning systems are achieving better performances at the cost of becoming increasingly complex. However, because of that, they become less interpretable, which may cause some distrust by the end-user of the system. This is especially important as these systems are pervasively being introduced to critical domains, such as the medical field. Representation Learning techniques are general methods for automatic feature computation. Nevertheless, these techniques are regarded as uninterpretable "black boxes". In this paper, we propose a methodology to enhance the interpretability of automatically extracted machine learning features. The proposed system is composed of a Restricted Boltzmann Machine for unsupervised feature learning, and a Random Forest classifier, which are combined to jointly consider existing correlations between imaging data, features, and target variables. We define two levels of interpretation: global and local. The former is devoted to understanding if the system learned the relevant relations in the data correctly, while the later is focused on predictions performed on a voxel- and patient-level. In addition, we propose a novel feature importance strategy that considers both imaging data and target variables, and we demonstrate the ability of the approach to leverage the interpretability of the obtained representation for the task at hand. We evaluated the proposed methodology in brain tumor segmentation and penumbra estimation in ischemic stroke lesions. We show the ability of the proposed methodology to unveil information regarding relationships between imaging modalities and extracted features and their usefulness for the task at hand. In both clinical scenarios, we demonstrate that the proposed methodology enhances the interpretability of automatically learned features, highlighting specific learning patterns that resemble how an expert extracts relevant data from medical images. Copyright © 2017 Elsevier B.V. All rights reserved.

Principal component analysis-based unsupervised feature extraction applied to in silico drug discovery for posttraumatic stress disorder-mediated heart disease.

PubMed

Taguchi, Y-h; Iwadate, Mitsuo; Umeyama, Hideaki

2015-04-30

Feature extraction (FE) is difficult, particularly if there are more features than samples, as small sample numbers often result in biased outcomes or overfitting. Furthermore, multiple sample classes often complicate FE because evaluating performance, which is usual in supervised FE, is generally harder than the two-class problem. Developing sample classification independent unsupervised methods would solve many of these problems. Two principal component analysis (PCA)-based FE, specifically, variational Bayes PCA (VBPCA) was extended to perform unsupervised FE, and together with conventional PCA (CPCA)-based unsupervised FE, were tested as sample classification independent unsupervised FE methods. VBPCA- and CPCA-based unsupervised FE both performed well when applied to simulated data, and a posttraumatic stress disorder (PTSD)-mediated heart disease data set that had multiple categorical class observations in mRNA/microRNA expression of stressed mouse heart. A critical set of PTSD miRNAs/mRNAs were identified that show aberrant expression between treatment and control samples, and significant, negative correlation with one another. Moreover, greater stability and biological feasibility than conventional supervised FE was also demonstrated. Based on the results obtained, in silico drug discovery was performed as translational validation of the methods. Our two proposed unsupervised FE methods (CPCA- and VBPCA-based) worked well on simulated data, and outperformed two conventional supervised FE methods on a real data set. Thus, these two methods have suggested equivalence for FE on categorical multiclass data sets, with potential translational utility for in silico drug discovery.

Multi-channel EEG-based sleep stage classification with joint collaborative representation and multiple kernel learning.

PubMed

Shi, Jun; Liu, Xiao; Li, Yan; Zhang, Qi; Li, Yingjie; Ying, Shihui

2015-10-30

Electroencephalography (EEG) based sleep staging is commonly used in clinical routine. Feature extraction and representation plays a crucial role in EEG-based automatic classification of sleep stages. Sparse representation (SR) is a state-of-the-art unsupervised feature learning method suitable for EEG feature representation. Collaborative representation (CR) is an effective data coding method used as a classifier. Here we use CR as a data representation method to learn features from the EEG signal. A joint collaboration model is established to develop a multi-view learning algorithm, and generate joint CR (JCR) codes to fuse and represent multi-channel EEG signals. A two-stage multi-view learning-based sleep staging framework is then constructed, in which JCR and joint sparse representation (JSR) algorithms first fuse and learning the feature representation from multi-channel EEG signals, respectively. Multi-view JCR and JSR features are then integrated and sleep stages recognized by a multiple kernel extreme learning machine (MK-ELM) algorithm with grid search. The proposed two-stage multi-view learning algorithm achieves superior performance for sleep staging. With a K-means clustering based dictionary, the mean classification accuracy, sensitivity and specificity are 81.10 ± 0.15%, 71.42 ± 0.66% and 94.57 ± 0.07%, respectively; while with the dictionary learned using the submodular optimization method, they are 80.29 ± 0.22%, 71.26 ± 0.78% and 94.38 ± 0.10%, respectively. The two-stage multi-view learning based sleep staging framework outperforms all other classification methods compared in this work, while JCR is superior to JSR. The proposed multi-view learning framework has the potential for sleep staging based on multi-channel or multi-modality polysomnography signals. Copyright © 2015 Elsevier B.V. All rights reserved.

True Zero-Training Brain-Computer Interfacing – An Online Study

PubMed Central

Kindermans, Pieter-Jan; Schreuder, Martijn; Schrauwen, Benjamin; Müller, Klaus-Robert; Tangermann, Michael

2014-01-01

Despite several approaches to realize subject-to-subject transfer of pre-trained classifiers, the full performance of a Brain-Computer Interface (BCI) for a novel user can only be reached by presenting the BCI system with data from the novel user. In typical state-of-the-art BCI systems with a supervised classifier, the labeled data is collected during a calibration recording, in which the user is asked to perform a specific task. Based on the known labels of this recording, the BCI's classifier can learn to decode the individual's brain signals. Unfortunately, this calibration recording consumes valuable time. Furthermore, it is unproductive with respect to the final BCI application, e.g. text entry. Therefore, the calibration period must be reduced to a minimum, which is especially important for patients with a limited concentration ability. The main contribution of this manuscript is an online study on unsupervised learning in an auditory event-related potential (ERP) paradigm. Our results demonstrate that the calibration recording can be bypassed by utilizing an unsupervised trained classifier, that is initialized randomly and updated during usage. Initially, the unsupervised classifier tends to make decoding mistakes, as the classifier might not have seen enough data to build a reliable model. Using a constant re-analysis of the previously spelled symbols, these initially misspelled symbols can be rectified posthoc when the classifier has learned to decode the signals. We compare the spelling performance of our unsupervised approach and of the unsupervised posthoc approach to the standard supervised calibration-based dogma for n = 10 healthy users. To assess the learning behavior of our approach, it is unsupervised trained from scratch three times per user. Even with the relatively low SNR of an auditory ERP paradigm, the results show that after a limited number of trials (30 trials), the unsupervised approach performs comparably to a classic supervised model. PMID:25068464

Optimized Structure of the Traffic Flow Forecasting Model With a Deep Learning Approach.

PubMed

Yang, Hao-Fan; Dillon, Tharam S; Chen, Yi-Ping Phoebe

2017-10-01

Forecasting accuracy is an important issue for successful intelligent traffic management, especially in the domain of traffic efficiency and congestion reduction. The dawning of the big data era brings opportunities to greatly improve prediction accuracy. In this paper, we propose a novel model, stacked autoencoder Levenberg-Marquardt model, which is a type of deep architecture of neural network approach aiming to improve forecasting accuracy. The proposed model is designed using the Taguchi method to develop an optimized structure and to learn traffic flow features through layer-by-layer feature granulation with a greedy layerwise unsupervised learning algorithm. It is applied to real-world data collected from the M6 freeway in the U.K. and is compared with three existing traffic predictors. To the best of our knowledge, this is the first time that an optimized structure of the traffic flow forecasting model with a deep learning approach is presented. The evaluation results demonstrate that the proposed model with an optimized structure has superior performance in traffic flow forecasting.

Learning Semantic Tags from Big Data for Clinical Text Representation.

PubMed

Li, Yanpeng; Liu, Hongfang

2015-01-01

In clinical text mining, it is one of the biggest challenges to represent medical terminologies and n-gram terms in sparse medical reports using either supervised or unsupervised methods. Addressing this issue, we propose a novel method for word and n-gram representation at semantic level. We first represent each word by its distance with a set of reference features calculated by reference distance estimator (RDE) learned from labeled and unlabeled data, and then generate new features using simple techniques of discretization, random sampling and merging. The new features are a set of binary rules that can be interpreted as semantic tags derived from word and n-grams. We show that the new features significantly outperform classical bag-of-words and n-grams in the task of heart disease risk factor extraction in i2b2 2014 challenge. It is promising to see that semantics tags can be used to replace the original text entirely with even better prediction performance as well as derive new rules beyond lexical level.

Clinical Named Entity Recognition Using Deep Learning Models.

PubMed

Wu, Yonghui; Jiang, Min; Xu, Jun; Zhi, Degui; Xu, Hua

2017-01-01

Clinical Named Entity Recognition (NER) is a critical natural language processing (NLP) task to extract important concepts (named entities) from clinical narratives. Researchers have extensively investigated machine learning models for clinical NER. Recently, there have been increasing efforts to apply deep learning models to improve the performance of current clinical NER systems. This study examined two popular deep learning architectures, the Convolutional Neural Network (CNN) and the Recurrent Neural Network (RNN), to extract concepts from clinical texts. We compared the two deep neural network architectures with three baseline Conditional Random Fields (CRFs) models and two state-of-the-art clinical NER systems using the i2b2 2010 clinical concept extraction corpus. The evaluation results showed that the RNN model trained with the word embeddings achieved a new state-of-the- art performance (a strict F1 score of 85.94%) for the defined clinical NER task, outperforming the best-reported system that used both manually defined and unsupervised learning features. This study demonstrates the advantage of using deep neural network architectures for clinical concept extraction, including distributed feature representation, automatic feature learning, and long-term dependencies capture. This is one of the first studies to compare the two widely used deep learning models and demonstrate the superior performance of the RNN model for clinical NER.

Clinical Named Entity Recognition Using Deep Learning Models

PubMed Central

Wu, Yonghui; Jiang, Min; Xu, Jun; Zhi, Degui; Xu, Hua

2017-01-01

Clinical Named Entity Recognition (NER) is a critical natural language processing (NLP) task to extract important concepts (named entities) from clinical narratives. Researchers have extensively investigated machine learning models for clinical NER. Recently, there have been increasing efforts to apply deep learning models to improve the performance of current clinical NER systems. This study examined two popular deep learning architectures, the Convolutional Neural Network (CNN) and the Recurrent Neural Network (RNN), to extract concepts from clinical texts. We compared the two deep neural network architectures with three baseline Conditional Random Fields (CRFs) models and two state-of-the-art clinical NER systems using the i2b2 2010 clinical concept extraction corpus. The evaluation results showed that the RNN model trained with the word embeddings achieved a new state-of-the- art performance (a strict F1 score of 85.94%) for the defined clinical NER task, outperforming the best-reported system that used both manually defined and unsupervised learning features. This study demonstrates the advantage of using deep neural network architectures for clinical concept extraction, including distributed feature representation, automatic feature learning, and long-term dependencies capture. This is one of the first studies to compare the two widely used deep learning models and demonstrate the superior performance of the RNN model for clinical NER. PMID:29854252

Supervised versus unsupervised categorization: two sides of the same coin?

PubMed

Pothos, Emmanuel M; Edwards, Darren J; Perlman, Amotz

2011-09-01

Supervised and unsupervised categorization have been studied in separate research traditions. A handful of studies have attempted to explore a possible convergence between the two. The present research builds on these studies, by comparing the unsupervised categorization results of Pothos et al. ( 2011 ; Pothos et al., 2008 ) with the results from two procedures of supervised categorization. In two experiments, we tested 375 participants with nine different stimulus sets and examined the relation between ease of learning of a classification, memory for a classification, and spontaneous preference for a classification. After taking into account the role of the number of category labels (clusters) in supervised learning, we found the three variables to be closely associated with each other. Our results provide encouragement for researchers seeking unified theoretical explanations for supervised and unsupervised categorization, but raise a range of challenging theoretical questions.

Penalized unsupervised learning with outliers

PubMed Central

Witten, Daniela M.

2013-01-01

We consider the problem of performing unsupervised learning in the presence of outliers – that is, observations that do not come from the same distribution as the rest of the data. It is known that in this setting, standard approaches for unsupervised learning can yield unsatisfactory results. For instance, in the presence of severe outliers, K-means clustering will often assign each outlier to its own cluster, or alternatively may yield distorted clusters in order to accommodate the outliers. In this paper, we take a new approach to extending existing unsupervised learning techniques to accommodate outliers. Our approach is an extension of a recent proposal for outlier detection in the regression setting. We allow each observation to take on an “error” term, and we penalize the errors using a group lasso penalty in order to encourage most of the observations’ errors to exactly equal zero. We show that this approach can be used in order to develop extensions of K-means clustering and principal components analysis that result in accurate outlier detection, as well as improved performance in the presence of outliers. These methods are illustrated in a simulation study and on two gene expression data sets, and connections with M-estimation are explored. PMID:23875057

An evaluation of unsupervised and supervised learning algorithms for clustering landscape types in the United States

USGS Publications Warehouse

Wendel, Jochen; Buttenfield, Barbara P.; Stanislawski, Larry V.

2016-01-01

Knowledge of landscape type can inform cartographic generalization of hydrographic features, because landscape characteristics provide an important geographic context that affects variation in channel geometry, flow pattern, and network configuration. Landscape types are characterized by expansive spatial gradients, lacking abrupt changes between adjacent classes; and as having a limited number of outliers that might confound classification. The US Geological Survey (USGS) is exploring methods to automate generalization of features in the National Hydrography Data set (NHD), to associate specific sequences of processing operations and parameters with specific landscape characteristics, thus obviating manual selection of a unique processing strategy for every NHD watershed unit. A chronology of methods to delineate physiographic regions for the United States is described, including a recent maximum likelihood classification based on seven input variables. This research compares unsupervised and supervised algorithms applied to these seven input variables, to evaluate and possibly refine the recent classification. Evaluation metrics for unsupervised methods include the Davies–Bouldin index, the Silhouette index, and the Dunn index as well as quantization and topographic error metrics. Cross validation and misclassification rate analysis are used to evaluate supervised classification methods. The paper reports the comparative analysis and its impact on the selection of landscape regions. The compared solutions show problems in areas of high landscape diversity. There is some indication that additional input variables, additional classes, or more sophisticated methods can refine the existing classification.

Learning Compact Binary Face Descriptor for Face Recognition.

PubMed

Lu, Jiwen; Liong, Venice Erin; Zhou, Xiuzhuang; Zhou, Jie

2015-10-01

Binary feature descriptors such as local binary patterns (LBP) and its variations have been widely used in many face recognition systems due to their excellent robustness and strong discriminative power. However, most existing binary face descriptors are hand-crafted, which require strong prior knowledge to engineer them by hand. In this paper, we propose a compact binary face descriptor (CBFD) feature learning method for face representation and recognition. Given each face image, we first extract pixel difference vectors (PDVs) in local patches by computing the difference between each pixel and its neighboring pixels. Then, we learn a feature mapping to project these pixel difference vectors into low-dimensional binary vectors in an unsupervised manner, where 1) the variance of all binary codes in the training set is maximized, 2) the loss between the original real-valued codes and the learned binary codes is minimized, and 3) binary codes evenly distribute at each learned bin, so that the redundancy information in PDVs is removed and compact binary codes are obtained. Lastly, we cluster and pool these binary codes into a histogram feature as the final representation for each face image. Moreover, we propose a coupled CBFD (C-CBFD) method by reducing the modality gap of heterogeneous faces at the feature level to make our method applicable to heterogeneous face recognition. Extensive experimental results on five widely used face datasets show that our methods outperform state-of-the-art face descriptors.

Unsupervised Pattern Classifier for Abnormality-Scaling of Vibration Features for Helicopter Gearbox Fault Diagnosis

NASA Technical Reports Server (NTRS)

Jammu, Vinay B.; Danai, Kourosh; Lewicki, David G.

1996-01-01

A new unsupervised pattern classifier is introduced for on-line detection of abnormality in features of vibration that are used for fault diagnosis of helicopter gearboxes. This classifier compares vibration features with their respective normal values and assigns them a value in (0, 1) to reflect their degree of abnormality. Therefore, the salient feature of this classifier is that it does not require feature values associated with faulty cases to identify abnormality. In order to cope with noise and changes in the operating conditions, an adaptation algorithm is incorporated that continually updates the normal values of the features. The proposed classifier is tested using experimental vibration features obtained from an OH-58A main rotor gearbox. The overall performance of this classifier is then evaluated by integrating the abnormality-scaled features for detection of faults. The fault detection results indicate that the performance of this classifier is comparable to the leading unsupervised neural networks: Kohonen's Feature Mapping and Adaptive Resonance Theory (AR72). This is significant considering that the independence of this classifier from fault-related features makes it uniquely suited to abnormality-scaling of vibration features for fault diagnosis.

Coexistence of Reward and Unsupervised Learning During the Operant Conditioning of Neural Firing Rates

PubMed Central

Kerr, Robert R.; Grayden, David B.; Thomas, Doreen A.; Gilson, Matthieu; Burkitt, Anthony N.

2014-01-01

A fundamental goal of neuroscience is to understand how cognitive processes, such as operant conditioning, are performed by the brain. Typical and well studied examples of operant conditioning, in which the firing rates of individual cortical neurons in monkeys are increased using rewards, provide an opportunity for insight into this. Studies of reward-modulated spike-timing-dependent plasticity (RSTDP), and of other models such as R-max, have reproduced this learning behavior, but they have assumed that no unsupervised learning is present (i.e., no learning occurs without, or independent of, rewards). We show that these models cannot elicit firing rate reinforcement while exhibiting both reward learning and ongoing, stable unsupervised learning. To fix this issue, we propose a new RSTDP model of synaptic plasticity based upon the observed effects that dopamine has on long-term potentiation and depression (LTP and LTD). We show, both analytically and through simulations, that our new model can exhibit unsupervised learning and lead to firing rate reinforcement. This requires that the strengthening of LTP by the reward signal is greater than the strengthening of LTD and that the reinforced neuron exhibits irregular firing. We show the robustness of our findings to spike-timing correlations, to the synaptic weight dependence that is assumed, and to changes in the mean reward. We also consider our model in the differential reinforcement of two nearby neurons. Our model aligns more strongly with experimental studies than previous models and makes testable predictions for future experiments. PMID:24475240

Classify epithelium-stroma in histopathological images based on deep transferable network.

PubMed

Yu, X; Zheng, H; Liu, C; Huang, Y; Ding, X

2018-04-20

Recently, the deep learning methods have received more attention in histopathological image analysis. However, the traditional deep learning methods assume that training data and test data have the same distributions, which causes certain limitations in real-world histopathological applications. However, it is costly to recollect a large amount of labeled histology data to train a new neural network for each specified image acquisition procedure even for similar tasks. In this paper, an unsupervised domain adaptation is introduced into a typical deep convolutional neural network (CNN) model to mitigate the repeating of the labels. The unsupervised domain adaptation is implemented by adding two regularisation terms, namely the feature-based adaptation and entropy minimisation, to the object function of a widely used CNN model called the AlexNet. Three independent public epithelium-stroma datasets were used to verify the proposed method. The experimental results have demonstrated that in the epithelium-stroma classification, the proposed method can achieve better performance than the commonly used deep learning methods and some existing deep domain adaptation methods. Therefore, the proposed method can be considered as a better option for the real-world applications of histopathological image analysis because there is no requirement for recollection of large-scale labeled data for every specified domain. © 2018 The Authors Journal of Microscopy © 2018 Royal Microscopical Society.

An Example of Unsupervised Networks Kohonen's Self-Organizing Feature Map

NASA Technical Reports Server (NTRS)

Niebur, Dagmar

1995-01-01

Kohonen's self-organizing feature map belongs to a class of unsupervised artificial neural network commonly referred to as topographic maps. It serves two purposes, the quantization and dimensionality reduction of date. A short description of its history and its biological context is given. We show that the inherent classification properties of the feature map make it a suitable candidate for solving the classification task in power system areas like load forecasting, fault diagnosis and security assessment.

Rough Set Based Splitting Criterion for Binary Decision Tree Classifiers

DTIC Science & Technology

2006-09-26

Alata O. Fernandez-Maloigne C., and Ferrie J.C. (2001). Unsupervised Algorithm for the Segmentation of Three-Dimensional Magnetic Resonance Brain ...instinctual and learned responses in the brain , causing it to make decisions based on patterns in the stimuli. Using this deceptively simple process...2001. [2] Bohn C. (1997). An Incremental Unsupervised Learning Scheme for Function Approximation. In: Proceedings of the 1997 IEEE International

An Introduction to Topic Modeling as an Unsupervised Machine Learning Way to Organize Text Information

ERIC Educational Resources Information Center

Snyder, Robin M.

2015-01-01

The field of topic modeling has become increasingly important over the past few years. Topic modeling is an unsupervised machine learning way to organize text (or image or DNA, etc.) information such that related pieces of text can be identified. This paper/session will present/discuss the current state of topic modeling, why it is important, and…

Instructional Videos for Unsupervised Harvesting and Learning of Action Examples

DTIC Science & Technology

2014-11-03

collection of image or video anno - tations has been tackled in different ways, but most existing methods still require a human in the loop. The...the views of ARO and NSF. 7. REFERENCES [1] C.-C. Chang and C.- J . Lin. LIBSVM: A library for support vector machines. In ACM Transactions on...feature encoding methods. In BMVC, 2011. [3] J . Chen, Y. Cui, G. Ye, D. Liu, and S.-F. Chang. Event-driven semantic concept discovery by exploiting

Robust location and spread measures for nonparametric probability density function estimation.

PubMed

López-Rubio, Ezequiel

2009-10-01

Robustness against outliers is a desirable property of any unsupervised learning scheme. In particular, probability density estimators benefit from incorporating this feature. A possible strategy to achieve this goal is to substitute the sample mean and the sample covariance matrix by more robust location and spread estimators. Here we use the L1-median to develop a nonparametric probability density function (PDF) estimator. We prove its most relevant properties, and we show its performance in density estimation and classification applications.

Neural networks for learning and prediction with applications to remote sensing and speech perception

NASA Astrophysics Data System (ADS)

Gjaja, Marin N.

1997-11-01

Neural networks for supervised and unsupervised learning are developed and applied to problems in remote sensing, continuous map learning, and speech perception. Adaptive Resonance Theory (ART) models are real-time neural networks for category learning, pattern recognition, and prediction. Unsupervised fuzzy ART networks synthesize fuzzy logic and neural networks, and supervised ARTMAP networks incorporate ART modules for prediction and classification. New ART and ARTMAP methods resulting from analyses of data structure, parameter specification, and category selection are developed. Architectural modifications providing flexibility for a variety of applications are also introduced and explored. A new methodology for automatic mapping from Landsat Thematic Mapper (TM) and terrain data, based on fuzzy ARTMAP, is developed. System capabilities are tested on a challenging remote sensing problem, prediction of vegetation classes in the Cleveland National Forest from spectral and terrain features. After training at the pixel level, performance is tested at the stand level, using sites not seen during training. Results are compared to those of maximum likelihood classifiers, back propagation neural networks, and K-nearest neighbor algorithms. Best performance is obtained using a hybrid system based on a convex combination of fuzzy ARTMAP and maximum likelihood predictions. This work forms the foundation for additional studies exploring fuzzy ARTMAP's capability to estimate class mixture composition for non-homogeneous sites. Exploratory simulations apply ARTMAP to the problem of learning continuous multidimensional mappings. A novel system architecture retains basic ARTMAP properties of incremental and fast learning in an on-line setting while adding components to solve this class of problems. The perceptual magnet effect is a language-specific phenomenon arising early in infant speech development that is characterized by a warping of speech sound perception. An unsupervised neural network model is proposed that embodies two principal hypotheses supported by experimental data--that sensory experience guides language-specific development of an auditory neural map and that a population vector can predict psychological phenomena based on map cell activities. Model simulations show how a nonuniform distribution of map cell firing preferences can develop from language-specific input and give rise to the magnet effect.

The Role of Architectural and Learning Constraints in Neural Network Models: A Case Study on Visual Space Coding.

PubMed

Testolin, Alberto; De Filippo De Grazia, Michele; Zorzi, Marco

2017-01-01

The recent "deep learning revolution" in artificial neural networks had strong impact and widespread deployment for engineering applications, but the use of deep learning for neurocomputational modeling has been so far limited. In this article we argue that unsupervised deep learning represents an important step forward for improving neurocomputational models of perception and cognition, because it emphasizes the role of generative learning as opposed to discriminative (supervised) learning. As a case study, we present a series of simulations investigating the emergence of neural coding of visual space for sensorimotor transformations. We compare different network architectures commonly used as building blocks for unsupervised deep learning by systematically testing the type of receptive fields and gain modulation developed by the hidden neurons. In particular, we compare Restricted Boltzmann Machines (RBMs), which are stochastic, generative networks with bidirectional connections trained using contrastive divergence, with autoencoders, which are deterministic networks trained using error backpropagation. For both learning architectures we also explore the role of sparse coding, which has been identified as a fundamental principle of neural computation. The unsupervised models are then compared with supervised, feed-forward networks that learn an explicit mapping between different spatial reference frames. Our simulations show that both architectural and learning constraints strongly influenced the emergent coding of visual space in terms of distribution of tuning functions at the level of single neurons. Unsupervised models, and particularly RBMs, were found to more closely adhere to neurophysiological data from single-cell recordings in the primate parietal cortex. These results provide new insights into how basic properties of artificial neural networks might be relevant for modeling neural information processing in biological systems.

The Role of Architectural and Learning Constraints in Neural Network Models: A Case Study on Visual Space Coding

PubMed Central

Testolin, Alberto; De Filippo De Grazia, Michele; Zorzi, Marco

2017-01-01

The recent “deep learning revolution” in artificial neural networks had strong impact and widespread deployment for engineering applications, but the use of deep learning for neurocomputational modeling has been so far limited. In this article we argue that unsupervised deep learning represents an important step forward for improving neurocomputational models of perception and cognition, because it emphasizes the role of generative learning as opposed to discriminative (supervised) learning. As a case study, we present a series of simulations investigating the emergence of neural coding of visual space for sensorimotor transformations. We compare different network architectures commonly used as building blocks for unsupervised deep learning by systematically testing the type of receptive fields and gain modulation developed by the hidden neurons. In particular, we compare Restricted Boltzmann Machines (RBMs), which are stochastic, generative networks with bidirectional connections trained using contrastive divergence, with autoencoders, which are deterministic networks trained using error backpropagation. For both learning architectures we also explore the role of sparse coding, which has been identified as a fundamental principle of neural computation. The unsupervised models are then compared with supervised, feed-forward networks that learn an explicit mapping between different spatial reference frames. Our simulations show that both architectural and learning constraints strongly influenced the emergent coding of visual space in terms of distribution of tuning functions at the level of single neurons. Unsupervised models, and particularly RBMs, were found to more closely adhere to neurophysiological data from single-cell recordings in the primate parietal cortex. These results provide new insights into how basic properties of artificial neural networks might be relevant for modeling neural information processing in biological systems. PMID:28377709

Deep SOMs for automated feature extraction and classification from big data streaming

NASA Astrophysics Data System (ADS)

Sakkari, Mohamed; Ejbali, Ridha; Zaied, Mourad

2017-03-01

In this paper, we proposed a deep self-organizing map model (Deep-SOMs) for automated features extracting and learning from big data streaming which we benefit from the framework Spark for real time streams and highly parallel data processing. The SOMs deep architecture is based on the notion of abstraction (patterns automatically extract from the raw data, from the less to more abstract). The proposed model consists of three hidden self-organizing layers, an input and an output layer. Each layer is made up of a multitude of SOMs, each map only focusing at local headmistress sub-region from the input image. Then, each layer trains the local information to generate more overall information in the higher layer. The proposed Deep-SOMs model is unique in terms of the layers architecture, the SOMs sampling method and learning. During the learning stage we use a set of unsupervised SOMs for feature extraction. We validate the effectiveness of our approach on large data sets such as Leukemia dataset and SRBCT. Results of comparison have shown that the Deep-SOMs model performs better than many existing algorithms for images classification.

Spike sorting based upon machine learning algorithms (SOMA).

PubMed

Horton, P M; Nicol, A U; Kendrick, K M; Feng, J F

2007-02-15

We have developed a spike sorting method, using a combination of various machine learning algorithms, to analyse electrophysiological data and automatically determine the number of sampled neurons from an individual electrode, and discriminate their activities. We discuss extensions to a standard unsupervised learning algorithm (Kohonen), as using a simple application of this technique would only identify a known number of clusters. Our extra techniques automatically identify the number of clusters within the dataset, and their sizes, thereby reducing the chance of misclassification. We also discuss a new pre-processing technique, which transforms the data into a higher dimensional feature space revealing separable clusters. Using principal component analysis (PCA) alone may not achieve this. Our new approach appends the features acquired using PCA with features describing the geometric shapes that constitute a spike waveform. To validate our new spike sorting approach, we have applied it to multi-electrode array datasets acquired from the rat olfactory bulb, and from the sheep infero-temporal cortex, and using simulated data. The SOMA sofware is available at http://www.sussex.ac.uk/Users/pmh20/spikes.

Color normalization of histology slides using graph regularized sparse NMF

NASA Astrophysics Data System (ADS)

Sha, Lingdao; Schonfeld, Dan; Sethi, Amit

2017-03-01

Computer based automatic medical image processing and quantification are becoming popular in digital pathology. However, preparation of histology slides can vary widely due to differences in staining equipment, procedures and reagents, which can reduce the accuracy of algorithms that analyze their color and texture information. To re- duce the unwanted color variations, various supervised and unsupervised color normalization methods have been proposed. Compared with supervised color normalization methods, unsupervised color normalization methods have advantages of time and cost efficient and universal applicability. Most of the unsupervised color normaliza- tion methods for histology are based on stain separation. Based on the fact that stain concentration cannot be negative and different parts of the tissue absorb different stains, nonnegative matrix factorization (NMF), and particular its sparse version (SNMF), are good candidates for stain separation. However, most of the existing unsupervised color normalization method like PCA, ICA, NMF and SNMF fail to consider important information about sparse manifolds that its pixels occupy, which could potentially result in loss of texture information during color normalization. Manifold learning methods like Graph Laplacian have proven to be very effective in interpreting high-dimensional data. In this paper, we propose a novel unsupervised stain separation method called graph regularized sparse nonnegative matrix factorization (GSNMF). By considering the sparse prior of stain concentration together with manifold information from high-dimensional image data, our method shows better performance in stain color deconvolution than existing unsupervised color deconvolution methods, especially in keeping connected texture information. To utilized the texture information, we construct a nearest neighbor graph between pixels within a spatial area of an image based on their distances using heat kernal in lαβ space. The representation of a pixel in the stain density space is constrained to follow the feature distance of the pixel to pixels in the neighborhood graph. Utilizing color matrix transfer method with the stain concentrations found using our GSNMF method, the color normalization performance was also better than existing methods.

Unsupervised learning in persistent sensing for target recognition by wireless ad hoc networks of ground-based sensors

NASA Astrophysics Data System (ADS)

Hortos, William S.

2008-04-01

In previous work by the author, effective persistent and pervasive sensing for recognition and tracking of battlefield targets were seen to be achieved, using intelligent algorithms implemented by distributed mobile agents over a composite system of unmanned aerial vehicles (UAVs) for persistence and a wireless network of unattended ground sensors for pervasive coverage of the mission environment. While simulated performance results for the supervised algorithms of the composite system are shown to provide satisfactory target recognition over relatively brief periods of system operation, this performance can degrade by as much as 50% as target dynamics in the environment evolve beyond the period of system operation in which the training data are representative. To overcome this limitation, this paper applies the distributed approach using mobile agents to the network of ground-based wireless sensors alone, without the UAV subsystem, to provide persistent as well as pervasive sensing for target recognition and tracking. The supervised algorithms used in the earlier work are supplanted by unsupervised routines, including competitive-learning neural networks (CLNNs) and new versions of support vector machines (SVMs) for characterization of an unknown target environment. To capture the same physical phenomena from battlefield targets as the composite system, the suite of ground-based sensors can be expanded to include imaging and video capabilities. The spatial density of deployed sensor nodes is increased to allow more precise ground-based location and tracking of detected targets by active nodes. The "swarm" mobile agents enabling WSN intelligence are organized in a three processing stages: detection, recognition and sustained tracking of ground targets. Features formed from the compressed sensor data are down-selected according to an information-theoretic algorithm that reduces redundancy within the feature set, reducing the dimension of samples used in the target recognition and tracking routines. Target tracking is based on simplified versions of Kalman filtration. Accuracy of recognition and tracking of implemented versions of the proposed suite of unsupervised algorithms is somewhat degraded from the ideal. Target recognition and tracking by supervised routines and by unsupervised SVM and CLNN routines in the ground-based WSN is evaluated in simulations using published system values and sensor data from vehicular targets in ground-surveillance scenarios. Results are compared with previously published performance for the system of the ground-based sensor network (GSN) and UAV swarm.

Deep generative learning for automated EHR diagnosis of traditional Chinese medicine.

PubMed

Liang, Zhaohui; Liu, Jun; Ou, Aihua; Zhang, Honglai; Li, Ziping; Huang, Jimmy Xiangji

2018-05-04

Computer-aided medical decision-making (CAMDM) is the method to utilize massive EMR data as both empirical and evidence support for the decision procedure of healthcare activities. Well-developed information infrastructure, such as hospital information systems and disease surveillance systems, provides abundant data for CAMDM. However, the complexity of EMR data with abstract medical knowledge makes the conventional model incompetent for the analysis. Thus a deep belief networks (DBN) based model is proposed to simulate the information analysis and decision-making procedure in medical practice. The purpose of this paper is to evaluate a deep learning architecture as an effective solution for CAMDM. A two-step model is applied in our study. At the first step, an optimized seven-layer deep belief network (DBN) is applied as an unsupervised learning algorithm to perform model training to acquire feature representation. Then a support vector machine model is adopted to DBN at the second step of the supervised learning. There are two data sets used in the experiments. One is a plain text data set indexed by medical experts. The other is a structured dataset on primary hypertension. The data are randomly divided to generate the training set for the unsupervised learning and the testing set for the supervised learning. The model performance is evaluated by the statistics of mean and variance, the average precision and coverage on the data sets. Two conventional shallow models (support vector machine / SVM and decision tree / DT) are applied as the comparisons to show the superiority of our proposed approach. The deep learning (DBN + SVM) model outperforms simple SVM and DT on two data sets in terms of all the evaluation measures, which confirms our motivation that the deep model is good at capturing the key features with less dependence when the index is built up by manpower. Our study shows the two-step deep learning model achieves high performance for medical information retrieval over the conventional shallow models. It is able to capture the features of both plain text and the highly-structured database of EMR data. The performance of the deep model is superior to the conventional shallow learning models such as SVM and DT. It is an appropriate knowledge-learning model for information retrieval of EMR system. Therefore, deep learning provides a good solution to improve the performance of CAMDM systems. Copyright © 2018. Published by Elsevier B.V.

Phonological Concept Learning.

PubMed

Moreton, Elliott; Pater, Joe; Pertsova, Katya

2017-01-01

Linguistic and non-linguistic pattern learning have been studied separately, but we argue for a comparative approach. Analogous inductive problems arise in phonological and visual pattern learning. Evidence from three experiments shows that human learners can solve them in analogous ways, and that human performance in both cases can be captured by the same models. We test GMECCS (Gradual Maximum Entropy with a Conjunctive Constraint Schema), an implementation of the Configural Cue Model (Gluck & Bower, ) in a Maximum Entropy phonotactic-learning framework (Goldwater & Johnson, ; Hayes & Wilson, ) with a single free parameter, against the alternative hypothesis that learners seek featurally simple algebraic rules ("rule-seeking"). We study the full typology of patterns introduced by Shepard, Hovland, and Jenkins () ("SHJ"), instantiated as both phonotactic patterns and visual analogs, using unsupervised training. Unlike SHJ, Experiments 1 and 2 found that both phonotactic and visual patterns that depended on fewer features could be more difficult than those that depended on more features, as predicted by GMECCS but not by rule-seeking. GMECCS also correctly predicted performance differences between stimulus subclasses within each pattern. A third experiment tried supervised training (which can facilitate rule-seeking in visual learning) to elicit simple rule-seeking phonotactic learning, but cue-based behavior persisted. We conclude that similar cue-based cognitive processes are available for phonological and visual concept learning, and hence that studying either kind of learning can lead to significant insights about the other. Copyright © 2015 Cognitive Science Society, Inc.

Recapitulation of Ayurveda constitution types by machine learning of phenotypic traits.

PubMed

Tiwari, Pradeep; Kutum, Rintu; Sethi, Tavpritesh; Shrivastava, Ankita; Girase, Bhushan; Aggarwal, Shilpi; Patil, Rutuja; Agarwal, Dhiraj; Gautam, Pramod; Agrawal, Anurag; Dash, Debasis; Ghosh, Saurabh; Juvekar, Sanjay; Mukerji, Mitali; Prasher, Bhavana

2017-01-01

In Ayurveda system of medicine individuals are classified into seven constitution types, "Prakriti", for assessing disease susceptibility and drug responsiveness. Prakriti evaluation involves clinical examination including questions about physiological and behavioural traits. A need was felt to develop models for accurately predicting Prakriti classes that have been shown to exhibit molecular differences. The present study was carried out on data of phenotypic attributes in 147 healthy individuals of three extreme Prakriti types, from a genetically homogeneous population of Western India. Unsupervised and supervised machine learning approaches were used to infer inherent structure of the data, and for feature selection and building classification models for Prakriti respectively. These models were validated in a North Indian population. Unsupervised clustering led to emergence of three natural clusters corresponding to three extreme Prakriti classes. The supervised modelling approaches could classify individuals, with distinct Prakriti types, in the training and validation sets. This study is the first to demonstrate that Prakriti types are distinct verifiable clusters within a multidimensional space of multiple interrelated phenotypic traits. It also provides a computational framework for predicting Prakriti classes from phenotypic attributes. This approach may be useful in precision medicine for stratification of endophenotypes in healthy and diseased populations.

Classification of damage in structural systems using time series analysis and supervised and unsupervised pattern recognition techniques

NASA Astrophysics Data System (ADS)

Omenzetter, Piotr; de Lautour, Oliver R.

2010-04-01

Developed for studying long, periodic records of various measured quantities, time series analysis methods are inherently suited and offer interesting possibilities for Structural Health Monitoring (SHM) applications. However, their use in SHM can still be regarded as an emerging application and deserves more studies. In this research, Autoregressive (AR) models were used to fit experimental acceleration time histories from two experimental structural systems, a 3- storey bookshelf-type laboratory structure and the ASCE Phase II SHM Benchmark Structure, in healthy and several damaged states. The coefficients of the AR models were chosen as damage sensitive features. Preliminary visual inspection of the large, multidimensional sets of AR coefficients to check the presence of clusters corresponding to different damage severities was achieved using Sammon mapping - an efficient nonlinear data compression technique. Systematic classification of damage into states based on the analysis of the AR coefficients was achieved using two supervised classification techniques: Nearest Neighbor Classification (NNC) and Learning Vector Quantization (LVQ), and one unsupervised technique: Self-organizing Maps (SOM). This paper discusses the performance of AR coefficients as damage sensitive features and compares the efficiency of the three classification techniques using experimental data.

Advanced soft computing diagnosis method for tumour grading.

PubMed

Papageorgiou, E I; Spyridonos, P P; Stylios, C D; Ravazoula, P; Groumpos, P P; Nikiforidis, G N

2006-01-01

To develop an advanced diagnostic method for urinary bladder tumour grading. A novel soft computing modelling methodology based on the augmentation of fuzzy cognitive maps (FCMs) with the unsupervised active Hebbian learning (AHL) algorithm is applied. One hundred and twenty-eight cases of urinary bladder cancer were retrieved from the archives of the Department of Histopathology, University Hospital of Patras, Greece. All tumours had been characterized according to the classical World Health Organization (WHO) grading system. To design the FCM model for tumour grading, three experts histopathologists defined the main histopathological features (concepts) and their impact on grade characterization. The resulted FCM model consisted of nine concepts. Eight concepts represented the main histopathological features for tumour grading. The ninth concept represented the tumour grade. To increase the classification ability of the FCM model, the AHL algorithm was applied to adjust the weights of the FCM. The proposed FCM grading model achieved a classification accuracy of 72.5%, 74.42% and 95.55% for tumours of grades I, II and III, respectively. An advanced computerized method to support tumour grade diagnosis decision was proposed and developed. The novelty of the method is based on employing the soft computing method of FCMs to represent specialized knowledge on histopathology and on augmenting FCMs ability using an unsupervised learning algorithm, the AHL. The proposed method performs with reasonably high accuracy compared to other existing methods and at the same time meets the physicians' requirements for transparency and explicability.

A new simple /spl infin/OH neuron model as a biologically plausible principal component analyzer.

PubMed

Jankovic, M V

2003-01-01

A new approach to unsupervised learning in a single-layer neural network is discussed. An algorithm for unsupervised learning based upon the Hebbian learning rule is presented. A simple neuron model is analyzed. A dynamic neural model, which contains both feed-forward and feedback connections between the input and the output, has been adopted. The, proposed learning algorithm could be more correctly named self-supervised rather than unsupervised. The solution proposed here is a modified Hebbian rule, in which the modification of the synaptic strength is proportional not to pre- and postsynaptic activity, but instead to the presynaptic and averaged value of postsynaptic activity. It is shown that the model neuron tends to extract the principal component from a stationary input vector sequence. Usually accepted additional decaying terms for the stabilization of the original Hebbian rule are avoided. Implementation of the basic Hebbian scheme would not lead to unrealistic growth of the synaptic strengths, thanks to the adopted network structure.

Deep Unsupervised Learning on a Desktop PC: A Primer for Cognitive Scientists.

PubMed

Testolin, Alberto; Stoianov, Ivilin; De Filippo De Grazia, Michele; Zorzi, Marco

2013-01-01

Deep belief networks hold great promise for the simulation of human cognition because they show how structured and abstract representations may emerge from probabilistic unsupervised learning. These networks build a hierarchy of progressively more complex distributed representations of the sensory data by fitting a hierarchical generative model. However, learning in deep networks typically requires big datasets and it can involve millions of connection weights, which implies that simulations on standard computers are unfeasible. Developing realistic, medium-to-large-scale learning models of cognition would therefore seem to require expertise in programing parallel-computing hardware, and this might explain why the use of this promising approach is still largely confined to the machine learning community. Here we show how simulations of deep unsupervised learning can be easily performed on a desktop PC by exploiting the processors of low cost graphic cards (graphic processor units) without any specific programing effort, thanks to the use of high-level programming routines (available in MATLAB or Python). We also show that even an entry-level graphic card can outperform a small high-performance computing cluster in terms of learning time and with no loss of learning quality. We therefore conclude that graphic card implementations pave the way for a widespread use of deep learning among cognitive scientists for modeling cognition and behavior.

Deep Unsupervised Learning on a Desktop PC: A Primer for Cognitive Scientists

PubMed Central

Testolin, Alberto; Stoianov, Ivilin; De Filippo De Grazia, Michele; Zorzi, Marco

2013-01-01

Deep belief networks hold great promise for the simulation of human cognition because they show how structured and abstract representations may emerge from probabilistic unsupervised learning. These networks build a hierarchy of progressively more complex distributed representations of the sensory data by fitting a hierarchical generative model. However, learning in deep networks typically requires big datasets and it can involve millions of connection weights, which implies that simulations on standard computers are unfeasible. Developing realistic, medium-to-large-scale learning models of cognition would therefore seem to require expertise in programing parallel-computing hardware, and this might explain why the use of this promising approach is still largely confined to the machine learning community. Here we show how simulations of deep unsupervised learning can be easily performed on a desktop PC by exploiting the processors of low cost graphic cards (graphic processor units) without any specific programing effort, thanks to the use of high-level programming routines (available in MATLAB or Python). We also show that even an entry-level graphic card can outperform a small high-performance computing cluster in terms of learning time and with no loss of learning quality. We therefore conclude that graphic card implementations pave the way for a widespread use of deep learning among cognitive scientists for modeling cognition and behavior. PMID:23653617

Audio-based, unsupervised machine learning reveals cyclic changes in earthquake mechanisms in the Geysers geothermal field, California

NASA Astrophysics Data System (ADS)

Holtzman, B. K.; Paté, A.; Paisley, J.; Waldhauser, F.; Repetto, D.; Boschi, L.

2017-12-01

The earthquake process reflects complex interactions of stress, fracture and frictional properties. New machine learning methods reveal patterns in time-dependent spectral properties of seismic signals and enable identification of changes in faulting processes. Our methods are based closely on those developed for music information retrieval and voice recognition, using the spectrogram instead of the waveform directly. Unsupervised learning involves identification of patterns based on differences among signals without any additional information provided to the algorithm. Clustering of 46,000 earthquakes of $0.3

Machine learning for neuroimaging with scikit-learn.

PubMed

Abraham, Alexandre; Pedregosa, Fabian; Eickenberg, Michael; Gervais, Philippe; Mueller, Andreas; Kossaifi, Jean; Gramfort, Alexandre; Thirion, Bertrand; Varoquaux, Gaël

2014-01-01

Statistical machine learning methods are increasingly used for neuroimaging data analysis. Their main virtue is their ability to model high-dimensional datasets, e.g., multivariate analysis of activation images or resting-state time series. Supervised learning is typically used in decoding or encoding settings to relate brain images to behavioral or clinical observations, while unsupervised learning can uncover hidden structures in sets of images (e.g., resting state functional MRI) or find sub-populations in large cohorts. By considering different functional neuroimaging applications, we illustrate how scikit-learn, a Python machine learning library, can be used to perform some key analysis steps. Scikit-learn contains a very large set of statistical learning algorithms, both supervised and unsupervised, and its application to neuroimaging data provides a versatile tool to study the brain.

Machine learning for neuroimaging with scikit-learn

PubMed Central

Abraham, Alexandre; Pedregosa, Fabian; Eickenberg, Michael; Gervais, Philippe; Mueller, Andreas; Kossaifi, Jean; Gramfort, Alexandre; Thirion, Bertrand; Varoquaux, Gaël

2014-01-01

Statistical machine learning methods are increasingly used for neuroimaging data analysis. Their main virtue is their ability to model high-dimensional datasets, e.g., multivariate analysis of activation images or resting-state time series. Supervised learning is typically used in decoding or encoding settings to relate brain images to behavioral or clinical observations, while unsupervised learning can uncover hidden structures in sets of images (e.g., resting state functional MRI) or find sub-populations in large cohorts. By considering different functional neuroimaging applications, we illustrate how scikit-learn, a Python machine learning library, can be used to perform some key analysis steps. Scikit-learn contains a very large set of statistical learning algorithms, both supervised and unsupervised, and its application to neuroimaging data provides a versatile tool to study the brain. PMID:24600388

IMMAN: free software for information theory-based chemometric analysis.

PubMed

Urias, Ricardo W Pino; Barigye, Stephen J; Marrero-Ponce, Yovani; García-Jacas, César R; Valdes-Martiní, José R; Perez-Gimenez, Facundo

2015-05-01

The features and theoretical background of a new and free computational program for chemometric analysis denominated IMMAN (acronym for Information theory-based CheMoMetrics ANalysis) are presented. This is multi-platform software developed in the Java programming language, designed with a remarkably user-friendly graphical interface for the computation of a collection of information-theoretic functions adapted for rank-based unsupervised and supervised feature selection tasks. A total of 20 feature selection parameters are presented, with the unsupervised and supervised frameworks represented by 10 approaches in each case. Several information-theoretic parameters traditionally used as molecular descriptors (MDs) are adapted for use as unsupervised rank-based feature selection methods. On the other hand, a generalization scheme for the previously defined differential Shannon's entropy is discussed, as well as the introduction of Jeffreys information measure for supervised feature selection. Moreover, well-known information-theoretic feature selection parameters, such as information gain, gain ratio, and symmetrical uncertainty are incorporated to the IMMAN software ( http://mobiosd-hub.com/imman-soft/ ), following an equal-interval discretization approach. IMMAN offers data pre-processing functionalities, such as missing values processing, dataset partitioning, and browsing. Moreover, single parameter or ensemble (multi-criteria) ranking options are provided. Consequently, this software is suitable for tasks like dimensionality reduction, feature ranking, as well as comparative diversity analysis of data matrices. Simple examples of applications performed with this program are presented. A comparative study between IMMAN and WEKA feature selection tools using the Arcene dataset was performed, demonstrating similar behavior. In addition, it is revealed that the use of IMMAN unsupervised feature selection methods improves the performance of both IMMAN and WEKA supervised algorithms. Graphic representation for Shannon's distribution of MD calculating software.

Nonlinear dimension reduction and clustering by Minimum Curvilinearity unfold neuropathic pain and tissue embryological classes.

PubMed

Cannistraci, Carlo Vittorio; Ravasi, Timothy; Montevecchi, Franco Maria; Ideker, Trey; Alessio, Massimo

2010-09-15

Nonlinear small datasets, which are characterized by low numbers of samples and very high numbers of measures, occur frequently in computational biology, and pose problems in their investigation. Unsupervised hybrid-two-phase (H2P) procedures-specifically dimension reduction (DR), coupled with clustering-provide valuable assistance, not only for unsupervised data classification, but also for visualization of the patterns hidden in high-dimensional feature space. 'Minimum Curvilinearity' (MC) is a principle that-for small datasets-suggests the approximation of curvilinear sample distances in the feature space by pair-wise distances over their minimum spanning tree (MST), and thus avoids the introduction of any tuning parameter. MC is used to design two novel forms of nonlinear machine learning (NML): Minimum Curvilinear embedding (MCE) for DR, and Minimum Curvilinear affinity propagation (MCAP) for clustering. Compared with several other unsupervised and supervised algorithms, MCE and MCAP, whether individually or combined in H2P, overcome the limits of classical approaches. High performance was attained in the visualization and classification of: (i) pain patients (proteomic measurements) in peripheral neuropathy; (ii) human organ tissues (genomic transcription factor measurements) on the basis of their embryological origin. MC provides a valuable framework to estimate nonlinear distances in small datasets. Its extension to large datasets is prefigured for novel NMLs. Classification of neuropathic pain by proteomic profiles offers new insights for future molecular and systems biology characterization of pain. Improvements in tissue embryological classification refine results obtained in an earlier study, and suggest a possible reinterpretation of skin attribution as mesodermal. https://sites.google.com/site/carlovittoriocannistraci/home.

Space coding for sensorimotor transformations can emerge through unsupervised learning.

PubMed

De Filippo De Grazia, Michele; Cutini, Simone; Lisi, Matteo; Zorzi, Marco

2012-08-01

The posterior parietal cortex (PPC) is fundamental for sensorimotor transformations because it combines multiple sensory inputs and posture signals into different spatial reference frames that drive motor programming. Here, we present a computational model mimicking the sensorimotor transformations occurring in the PPC. A recurrent neural network with one layer of hidden neurons (restricted Boltzmann machine) learned a stochastic generative model of the sensory data without supervision. After the unsupervised learning phase, the activity of the hidden neurons was used to compute a motor program (a population code on a bidimensional map) through a simple linear projection and delta rule learning. The average motor error, calculated as the difference between the expected and the computed output, was less than 3°. Importantly, analyses of the hidden neurons revealed gain-modulated visual receptive fields, thereby showing that space coding for sensorimotor transformations similar to that observed in the PPC can emerge through unsupervised learning. These results suggest that gain modulation is an efficient coding strategy to integrate visual and postural information toward the generation of motor commands.

Identifying quantum phase transitions with adversarial neural networks

NASA Astrophysics Data System (ADS)

Huembeli, Patrick; Dauphin, Alexandre; Wittek, Peter

2018-04-01

The identification of phases of matter is a challenging task, especially in quantum mechanics, where the complexity of the ground state appears to grow exponentially with the size of the system. Traditionally, physicists have to identify the relevant order parameters for the classification of the different phases. We here follow a radically different approach: we address this problem with a state-of-the-art deep learning technique, adversarial domain adaptation. We derive the phase diagram of the whole parameter space starting from a fixed and known subspace using unsupervised learning. This method has the advantage that the input of the algorithm can be directly the ground state without any ad hoc feature engineering. Furthermore, the dimension of the parameter space is unrestricted. More specifically, the input data set contains both labeled and unlabeled data instances. The first kind is a system that admits an accurate analytical or numerical solution, and one can recover its phase diagram. The second type is the physical system with an unknown phase diagram. Adversarial domain adaptation uses both types of data to create invariant feature extracting layers in a deep learning architecture. Once these layers are trained, we can attach an unsupervised learner to the network to find phase transitions. We show the success of this technique by applying it on several paradigmatic models: the Ising model with different temperatures, the Bose-Hubbard model, and the Su-Schrieffer-Heeger model with disorder. The method finds unknown transitions successfully and predicts transition points in close agreement with standard methods. This study opens the door to the classification of physical systems where the phase boundaries are complex such as the many-body localization problem or the Bose glass phase.

Nonequilibrium thermodynamics of restricted Boltzmann machines.

PubMed

Salazar, Domingos S P

2017-08-01

In this work, we analyze the nonequilibrium thermodynamics of a class of neural networks known as restricted Boltzmann machines (RBMs) in the context of unsupervised learning. We show how the network is described as a discrete Markov process and how the detailed balance condition and the Maxwell-Boltzmann equilibrium distribution are sufficient conditions for a complete thermodynamics description, including nonequilibrium fluctuation theorems. Numerical simulations in a fully trained RBM are performed and the heat exchange fluctuation theorem is verified with excellent agreement to the theory. We observe how the contrastive divergence functional, mostly used in unsupervised learning of RBMs, is closely related to nonequilibrium thermodynamic quantities. We also use the framework to interpret the estimation of the partition function of RBMs with the annealed importance sampling method from a thermodynamics standpoint. Finally, we argue that unsupervised learning of RBMs is equivalent to a work protocol in a system driven by the laws of thermodynamics in the absence of labeled data.

Quantum-Enhanced Machine Learning

NASA Astrophysics Data System (ADS)

Dunjko, Vedran; Taylor, Jacob M.; Briegel, Hans J.

2016-09-01

The emerging field of quantum machine learning has the potential to substantially aid in the problems and scope of artificial intelligence. This is only enhanced by recent successes in the field of classical machine learning. In this work we propose an approach for the systematic treatment of machine learning, from the perspective of quantum information. Our approach is general and covers all three main branches of machine learning: supervised, unsupervised, and reinforcement learning. While quantum improvements in supervised and unsupervised learning have been reported, reinforcement learning has received much less attention. Within our approach, we tackle the problem of quantum enhancements in reinforcement learning as well, and propose a systematic scheme for providing improvements. As an example, we show that quadratic improvements in learning efficiency, and exponential improvements in performance over limited time periods, can be obtained for a broad class of learning problems.

Artificial Intelligence in Cardiology.

PubMed

Johnson, Kipp W; Torres Soto, Jessica; Glicksberg, Benjamin S; Shameer, Khader; Miotto, Riccardo; Ali, Mohsin; Ashley, Euan; Dudley, Joel T

2018-06-12

Artificial intelligence and machine learning are poised to influence nearly every aspect of the human condition, and cardiology is not an exception to this trend. This paper provides a guide for clinicians on relevant aspects of artificial intelligence and machine learning, reviews selected applications of these methods in cardiology to date, and identifies how cardiovascular medicine could incorporate artificial intelligence in the future. In particular, the paper first reviews predictive modeling concepts relevant to cardiology such as feature selection and frequent pitfalls such as improper dichotomization. Second, it discusses common algorithms used in supervised learning and reviews selected applications in cardiology and related disciplines. Third, it describes the advent of deep learning and related methods collectively called unsupervised learning, provides contextual examples both in general medicine and in cardiovascular medicine, and then explains how these methods could be applied to enable precision cardiology and improve patient outcomes. Copyright © 2018 The Authors. Published by Elsevier Inc. All rights reserved.

Residential roof condition assessment system using deep learning

NASA Astrophysics Data System (ADS)

Wang, Fan; Kerekes, John P.; Xu, Zhuoyi; Wang, Yandong

2018-01-01

The emergence of high resolution (HR) and ultra high resolution (UHR) airborne remote sensing imagery is enabling humans to move beyond traditional land cover analysis applications to the detailed characterization of surface objects. A residential roof condition assessment method using techniques from deep learning is presented. The proposed method operates on individual roofs and divides the task into two stages: (1) roof segmentation, followed by (2) condition classification of the segmented roof regions. As the first step in this process, a self-tuning method is proposed to segment the images into small homogeneous areas. The segmentation is initialized with simple linear iterative clustering followed by deep learned feature extraction and region merging, with the optimal result selected by an unsupervised index, Q. After the segmentation, a pretrained residual network is fine-tuned on the augmented roof segments using a proposed k-pixel extension technique for classification. The effectiveness of the proposed algorithm was demonstrated on both HR and UHR imagery collected by EagleView over different study sites. The proposed algorithm has yielded promising results and has outperformed traditional machine learning methods using hand-crafted features.

SUSTAIN: a network model of category learning.

PubMed

Love, Bradley C; Medin, Douglas L; Gureckis, Todd M

2004-04-01

SUSTAIN (Supervised and Unsupervised STratified Adaptive Incremental Network) is a model of how humans learn categories from examples. SUSTAIN initially assumes a simple category structure. If simple solutions prove inadequate and SUSTAIN is confronted with a surprising event (e.g., it is told that a bat is a mammal instead of a bird), SUSTAIN recruits an additional cluster to represent the surprising event. Newly recruited clusters are available to explain future events and can themselves evolve into prototypes-attractors-rules. SUSTAIN's discovery of category substructure is affected not only by the structure of the world but by the nature of the learning task and the learner's goals. SUSTAIN successfully extends category learning models to studies of inference learning, unsupervised learning, category construction, and contexts in which identification learning is faster than classification learning.

A novel method for intelligent fault diagnosis of rolling bearings using ensemble deep auto-encoders

NASA Astrophysics Data System (ADS)

Shao, Haidong; Jiang, Hongkai; Lin, Ying; Li, Xingqiu

2018-03-01

Automatic and accurate identification of rolling bearings fault categories, especially for the fault severities and fault orientations, is still a major challenge in rotating machinery fault diagnosis. In this paper, a novel method called ensemble deep auto-encoders (EDAEs) is proposed for intelligent fault diagnosis of rolling bearings. Firstly, different activation functions are employed as the hidden functions to design a series of auto-encoders (AEs) with different characteristics. Secondly, EDAEs are constructed with various auto-encoders for unsupervised feature learning from the measured vibration signals. Finally, a combination strategy is designed to ensure accurate and stable diagnosis results. The proposed method is applied to analyze the experimental bearing vibration signals. The results confirm that the proposed method can get rid of the dependence on manual feature extraction and overcome the limitations of individual deep learning models, which is more effective than the existing intelligent diagnosis methods.

Comprehensive machine learning analysis of Hydra behavior reveals a stable basal behavioral repertoire

PubMed Central

Taralova, Ekaterina; Dupre, Christophe; Yuste, Rafael

2018-01-01

Animal behavior has been studied for centuries, but few efficient methods are available to automatically identify and classify it. Quantitative behavioral studies have been hindered by the subjective and imprecise nature of human observation, and the slow speed of annotating behavioral data. Here, we developed an automatic behavior analysis pipeline for the cnidarian Hydra vulgaris using machine learning. We imaged freely behaving Hydra, extracted motion and shape features from the videos, and constructed a dictionary of visual features to classify pre-defined behaviors. We also identified unannotated behaviors with unsupervised methods. Using this analysis pipeline, we quantified 6 basic behaviors and found surprisingly similar behavior statistics across animals within the same species, regardless of experimental conditions. Our analysis indicates that the fundamental behavioral repertoire of Hydra is stable. This robustness could reflect a homeostatic neural control of "housekeeping" behaviors which could have been already present in the earliest nervous systems. PMID:29589829

EEG-based driver fatigue detection using hybrid deep generic model.

PubMed

Phyo Phyo San; Sai Ho Ling; Rifai Chai; Tran, Yvonne; Craig, Ashley; Hung Nguyen

2016-08-01

Classification of electroencephalography (EEG)-based application is one of the important process for biomedical engineering. Driver fatigue is a major case of traffic accidents worldwide and considered as a significant problem in recent decades. In this paper, a hybrid deep generic model (DGM)-based support vector machine is proposed for accurate detection of driver fatigue. Traditionally, a probabilistic DGM with deep architecture is quite good at learning invariant features, but it is not always optimal for classification due to its trainable parameters are in the middle layer. Alternatively, Support Vector Machine (SVM) itself is unable to learn complicated invariance, but produces good decision surface when applied to well-behaved features. Consolidating unsupervised high-level feature extraction techniques, DGM and SVM classification makes the integrated framework stronger and enhance mutually in feature extraction and classification. The experimental results showed that the proposed DBN-based driver fatigue monitoring system achieves better testing accuracy of 73.29 % with 91.10 % sensitivity and 55.48 % specificity. In short, the proposed hybrid DGM-based SVM is an effective method for the detection of driver fatigue in EEG.

Discovering motion primitives for unsupervised grouping and one-shot learning of human actions, gestures, and expressions.

PubMed

Yang, Yang; Saleemi, Imran; Shah, Mubarak

2013-07-01

This paper proposes a novel representation of articulated human actions and gestures and facial expressions. The main goals of the proposed approach are: 1) to enable recognition using very few examples, i.e., one or k-shot learning, and 2) meaningful organization of unlabeled datasets by unsupervised clustering. Our proposed representation is obtained by automatically discovering high-level subactions or motion primitives, by hierarchical clustering of observed optical flow in four-dimensional, spatial, and motion flow space. The completely unsupervised proposed method, in contrast to state-of-the-art representations like bag of video words, provides a meaningful representation conducive to visual interpretation and textual labeling. Each primitive action depicts an atomic subaction, like directional motion of limb or torso, and is represented by a mixture of four-dimensional Gaussian distributions. For one--shot and k-shot learning, the sequence of primitive labels discovered in a test video are labeled using KL divergence, and can then be represented as a string and matched against similar strings of training videos. The same sequence can also be collapsed into a histogram of primitives or be used to learn a Hidden Markov model to represent classes. We have performed extensive experiments on recognition by one and k-shot learning as well as unsupervised action clustering on six human actions and gesture datasets, a composite dataset, and a database of facial expressions. These experiments confirm the validity and discriminative nature of the proposed representation.

Unsupervised heart-rate estimation in wearables with Liquid states and a probabilistic readout.

PubMed

Das, Anup; Pradhapan, Paruthi; Groenendaal, Willemijn; Adiraju, Prathyusha; Rajan, Raj Thilak; Catthoor, Francky; Schaafsma, Siebren; Krichmar, Jeffrey L; Dutt, Nikil; Van Hoof, Chris

2018-03-01

Heart-rate estimation is a fundamental feature of modern wearable devices. In this paper we propose a machine learning technique to estimate heart-rate from electrocardiogram (ECG) data collected using wearable devices. The novelty of our approach lies in (1) encoding spatio-temporal properties of ECG signals directly into spike train and using this to excite recurrently connected spiking neurons in a Liquid State Machine computation model; (2) a novel learning algorithm; and (3) an intelligently designed unsupervised readout based on Fuzzy c-Means clustering of spike responses from a subset of neurons (Liquid states), selected using particle swarm optimization. Our approach differs from existing works by learning directly from ECG signals (allowing personalization), without requiring costly data annotations. Additionally, our approach can be easily implemented on state-of-the-art spiking-based neuromorphic systems, offering high accuracy, yet significantly low energy footprint, leading to an extended battery-life of wearable devices. We validated our approach with CARLsim, a GPU accelerated spiking neural network simulator modeling Izhikevich spiking neurons with Spike Timing Dependent Plasticity (STDP) and homeostatic scaling. A range of subjects is considered from in-house clinical trials and public ECG databases. Results show high accuracy and low energy footprint in heart-rate estimation across subjects with and without cardiac irregularities, signifying the strong potential of this approach to be integrated in future wearable devices. Copyright © 2018 Elsevier Ltd. All rights reserved.

Data Exploration using Unsupervised Feature Extraction for Mixed Micro-Seismic Signals

NASA Astrophysics Data System (ADS)

Meyer, Matthias; Weber, Samuel; Beutel, Jan

2017-04-01

We present a system for the analysis of data originating in a multi-sensor and multi-year experiment focusing on slope stability and its underlying processes in fractured permafrost rock walls undertaken at 3500m a.s.l. on the Matterhorn Hörnligrat, (Zermatt, Switzerland). This system incorporates facilities for the transmission, management and storage of large-scales of data ( 7 GB/day), preprocessing and aggregation of multiple sensor types, machine-learning based automatic feature extraction for micro-seismic and acoustic emission data and interactive web-based visualization of the data. Specifically, a combination of three types of sensors are used to profile the frequency spectrum from 1 Hz to 80 kHz with the goal to identify the relevant destructive processes (e.g. micro-cracking and fracture propagation) leading to the eventual destabilization of large rock masses. The sensors installed for this profiling experiment (2 geophones, 1 accelerometers and 2 piezo-electric sensors for detecting acoustic emission), are further augmented with sensors originating from a previous activity focusing on long-term monitoring of temperature evolution and rock kinematics with the help of wireless sensor networks (crackmeters, cameras, weather station, rock temperature profiles, differential GPS) [Hasler2012]. In raw format, the data generated by the different types of sensors, specifically the micro-seismic and acoustic emission sensors, is strongly heterogeneous, in part unsynchronized and the storage and processing demand is large. Therefore, a purpose-built signal preprocessing and event-detection system is used. While the analysis of data from each individual sensor follows established methods, the application of all these sensor types in combination within a field experiment is unique. Furthermore, experience and methods from using such sensors in laboratory settings cannot be readily transferred to the mountain field site setting with its scale and full exposure to the natural environment. Consequently, many state-of-the-art algorithms for big data analysis and event classification requiring a ground truth dataset cannot be applied. The above mentioned challenges require a tool for data exploration. In the presented system, data exploration is supported by unsupervised feature learning based on convolutional neural networks, which is used to automatically extract common features for preliminary clustering and outlier detection. With this information, an interactive web-tool allows for a fast identification of interesting time segments on which segment-selective algorithms for visualization, feature extraction and statistics can be applied. The combination of manual labeling based and unsupervised feature extraction provides an event catalog for classification of different characteristic events related to internal progression of micro-crack in steep fractured bedrock permafrost. References Hasler, A., S. Gruber, and J. Beutel (2012), Kinematics of steep bedrock permafrost, J. Geophys. Res., 117, F01016, doi:10.1029/2011JF001981.

Land use and land cover classification for rural residential areas in China using soft-probability cascading of multifeatures

NASA Astrophysics Data System (ADS)

Zhang, Bin; Liu, Yueyan; Zhang, Zuyu; Shen, Yonglin

2017-10-01

A multifeature soft-probability cascading scheme to solve the problem of land use and land cover (LULC) classification using high-spatial-resolution images to map rural residential areas in China is proposed. The proposed method is used to build midlevel LULC features. Local features are frequently considered as low-level feature descriptors in a midlevel feature learning method. However, spectral and textural features, which are very effective low-level features, are neglected. The acquisition of the dictionary of sparse coding is unsupervised, and this phenomenon reduces the discriminative power of the midlevel feature. Thus, we propose to learn supervised features based on sparse coding, a support vector machine (SVM) classifier, and a conditional random field (CRF) model to utilize the different effective low-level features and improve the discriminability of midlevel feature descriptors. First, three kinds of typical low-level features, namely, dense scale-invariant feature transform, gray-level co-occurrence matrix, and spectral features, are extracted separately. Second, combined with sparse coding and the SVM classifier, the probabilities of the different LULC classes are inferred to build supervised feature descriptors. Finally, the CRF model, which consists of two parts: unary potential and pairwise potential, is employed to construct an LULC classification map. Experimental results show that the proposed classification scheme can achieve impressive performance when the total accuracy reached about 87%.

Information processing of motion in facial expression and the geometry of dynamical systems

NASA Astrophysics Data System (ADS)

Assadi, Amir H.; Eghbalnia, Hamid; McMenamin, Brenton W.

2005-01-01

An interesting problem in analysis of video data concerns design of algorithms that detect perceptually significant features in an unsupervised manner, for instance methods of machine learning for automatic classification of human expression. A geometric formulation of this genre of problems could be modeled with help of perceptual psychology. In this article, we outline one approach for a special case where video segments are to be classified according to expression of emotion or other similar facial motions. The encoding of realistic facial motions that convey expression of emotions for a particular person P forms a parameter space XP whose study reveals the "objective geometry" for the problem of unsupervised feature detection from video. The geometric features and discrete representation of the space XP are independent of subjective evaluations by observers. While the "subjective geometry" of XP varies from observer to observer, levels of sensitivity and variation in perception of facial expressions appear to share a certain level of universality among members of similar cultures. Therefore, statistical geometry of invariants of XP for a sample of population could provide effective algorithms for extraction of such features. In cases where frequency of events is sufficiently large in the sample data, a suitable framework could be provided to facilitate the information-theoretic organization and study of statistical invariants of such features. This article provides a general approach to encode motion in terms of a particular genre of dynamical systems and the geometry of their flow. An example is provided to illustrate the general theory.

A graph lattice approach to maintaining and learning dense collections of subgraphs as image features.

PubMed

Saund, Eric

2013-10-01

Effective object and scene classification and indexing depend on extraction of informative image features. This paper shows how large families of complex image features in the form of subgraphs can be built out of simpler ones through construction of a graph lattice—a hierarchy of related subgraphs linked in a lattice. Robustness is achieved by matching many overlapping and redundant subgraphs, which allows the use of inexpensive exact graph matching, instead of relying on expensive error-tolerant graph matching to a minimal set of ideal model graphs. Efficiency in exact matching is gained by exploitation of the graph lattice data structure. Additionally, the graph lattice enables methods for adaptively growing a feature space of subgraphs tailored to observed data. We develop the approach in the domain of rectilinear line art, specifically for the practical problem of document forms recognition. We are especially interested in methods that require only one or very few labeled training examples per category. We demonstrate two approaches to using the subgraph features for this purpose. Using a bag-of-words feature vector we achieve essentially single-instance learning on a benchmark forms database, following an unsupervised clustering stage. Further performance gains are achieved on a more difficult dataset using a feature voting method and feature selection procedure.

e-IQ and IQ knowledge mining for generalized LDA

NASA Astrophysics Data System (ADS)

Jenkins, Jeffrey; van Bergem, Rutger; Sweet, Charles; Vietsch, Eveline; Szu, Harold

2015-05-01

How can the human brain uncover patterns, associations and features in real-time, real-world data? There must be a general strategy used to transform raw signals into useful features, but representing this generalization in the context of our information extraction tool set is lacking. In contrast to Big Data (BD), Large Data Analysis (LDA) has become a reachable multi-disciplinary goal in recent years due in part to high performance computers and algorithm development, as well as the availability of large data sets. However, the experience of Machine Learning (ML) and information communities has not been generalized into an intuitive framework that is useful to researchers across disciplines. The data exploration phase of data mining is a prime example of this unspoken, ad-hoc nature of ML - the Computer Scientist works with a Subject Matter Expert (SME) to understand the data, and then build tools (i.e. classifiers, etc.) which can benefit the SME and the rest of the researchers in that field. We ask, why is there not a tool to represent information in a meaningful way to the researcher asking the question? Meaning is subjective and contextual across disciplines, so to ensure robustness, we draw examples from several disciplines and propose a generalized LDA framework for independent data understanding of heterogeneous sources which contribute to Knowledge Discovery in Databases (KDD). Then, we explore the concept of adaptive Information resolution through a 6W unsupervised learning methodology feedback system. In this paper, we will describe the general process of man-machine interaction in terms of an asymmetric directed graph theory (digging for embedded knowledge), and model the inverse machine-man feedback (digging for tacit knowledge) as an ANN unsupervised learning methodology. Finally, we propose a collective learning framework which utilizes a 6W semantic topology to organize heterogeneous knowledge and diffuse information to entities within a society in a personalized way.

Model–Free Visualization of Suspicious Lesions in Breast MRI Based on Supervised and Unsupervised Learning

PubMed Central

Twellmann, Thorsten; Meyer-Baese, Anke; Lange, Oliver; Foo, Simon; Nattkemper, Tim W.

2008-01-01

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) has become an important tool in breast cancer diagnosis, but evaluation of multitemporal 3D image data holds new challenges for human observers. To aid the image analysis process, we apply supervised and unsupervised pattern recognition techniques for computing enhanced visualizations of suspicious lesions in breast MRI data. These techniques represent an important component of future sophisticated computer-aided diagnosis (CAD) systems and support the visual exploration of spatial and temporal features of DCE-MRI data stemming from patients with confirmed lesion diagnosis. By taking into account the heterogeneity of cancerous tissue, these techniques reveal signals with malignant, benign and normal kinetics. They also provide a regional subclassification of pathological breast tissue, which is the basis for pseudo-color presentations of the image data. Intelligent medical systems are expected to have substantial implications in healthcare politics by contributing to the diagnosis of indeterminate breast lesions by non-invasive imaging. PMID:19255616

Informatics and machine learning to define the phenotype.

PubMed

Basile, Anna Okula; Ritchie, Marylyn DeRiggi

2018-03-01

For the past decade, the focus of complex disease research has been the genotype. From technological advancements to the development of analysis methods, great progress has been made. However, advances in our definition of the phenotype have remained stagnant. Phenotype characterization has recently emerged as an exciting area of informatics and machine learning. The copious amounts of diverse biomedical data that have been collected may be leveraged with data-driven approaches to elucidate trait-related features and patterns. Areas covered: In this review, the authors discuss the phenotype in traditional genetic associations and the challenges this has imposed.Approaches for phenotype refinement that can aid in more accurate characterization of traits are also discussed. Further, the authors highlight promising machine learning approaches for establishing a phenotype and the challenges of electronic health record (EHR)-derived data. Expert commentary: The authors hypothesize that through unsupervised machine learning, data-driven approaches can be used to define phenotypes rather than relying on expert clinician knowledge. Through the use of machine learning and an unbiased set of features extracted from clinical repositories, researchers will have the potential to further understand complex traits and identify patient subgroups. This knowledge may lead to more preventative and precise clinical care.

A unified framework for automatic wound segmentation and analysis with deep convolutional neural networks.

PubMed

Wang, Changhan; Yan, Xinchen; Smith, Max; Kochhar, Kanika; Rubin, Marcie; Warren, Stephen M; Wrobel, James; Lee, Honglak

2015-01-01

Wound surface area changes over multiple weeks are highly predictive of the wound healing process. Furthermore, the quality and quantity of the tissue in the wound bed also offer important prognostic information. Unfortunately, accurate measurements of wound surface area changes are out of reach in the busy wound practice setting. Currently, clinicians estimate wound size by estimating wound width and length using a scalpel after wound treatment, which is highly inaccurate. To address this problem, we propose an integrated system to automatically segment wound regions and analyze wound conditions in wound images. Different from previous segmentation techniques which rely on handcrafted features or unsupervised approaches, our proposed deep learning method jointly learns task-relevant visual features and performs wound segmentation. Moreover, learned features are applied to further analysis of wounds in two ways: infection detection and healing progress prediction. To the best of our knowledge, this is the first attempt to automate long-term predictions of general wound healing progress. Our method is computationally efficient and takes less than 5 seconds per wound image (480 by 640 pixels) on a typical laptop computer. Our evaluations on a large-scale wound database demonstrate the effectiveness and reliability of the proposed system.

Complex scenes and situations visualization in hierarchical learning algorithm with dynamic 3D NeoAxis engine

NASA Astrophysics Data System (ADS)

Graham, James; Ternovskiy, Igor V.

2013-06-01

We applied a two stage unsupervised hierarchical learning system to model complex dynamic surveillance and cyber space monitoring systems using a non-commercial version of the NeoAxis visualization software. The hierarchical scene learning and recognition approach is based on hierarchical expectation maximization, and was linked to a 3D graphics engine for validation of learning and classification results and understanding the human - autonomous system relationship. Scene recognition is performed by taking synthetically generated data and feeding it to a dynamic logic algorithm. The algorithm performs hierarchical recognition of the scene by first examining the features of the objects to determine which objects are present, and then determines the scene based on the objects present. This paper presents a framework within which low level data linked to higher-level visualization can provide support to a human operator and be evaluated in a detailed and systematic way.

Classification-free threat detection based on material-science-informed clustering

NASA Astrophysics Data System (ADS)

Yuan, Siyang; Wolter, Scott D.; Greenberg, Joel A.

2017-05-01

X-ray diffraction (XRD) is well-known for yielding composition and structural information about a material. However, in some applications (such as threat detection in aviation security), the properties of a material are more relevant to the task than is a detailed material characterization. Furthermore, the requirement that one first identify a material before determining its class may be difficult or even impossible for a sufficiently large pool of potentially present materials. We therefore seek to learn relevant composition-structure-property relationships between materials to enable material-identification-free classification. We use an expert-informed, data-driven approach operating on a library of XRD spectra from a broad array of stream of commerce materials. We investigate unsupervised learning techniques in order to learn about naturally emergent groupings, and apply supervised learning techniques to determine how well XRD features can be used to separate user-specified classes in the presence of different types and degrees of signal degradation.

Deep Learning Representation from Electroencephalography of Early-Stage Creutzfeldt-Jakob Disease and Features for Differentiation from Rapidly Progressive Dementia.

PubMed

Morabito, Francesco Carlo; Campolo, Maurizio; Mammone, Nadia; Versaci, Mario; Franceschetti, Silvana; Tagliavini, Fabrizio; Sofia, Vito; Fatuzzo, Daniela; Gambardella, Antonio; Labate, Angelo; Mumoli, Laura; Tripodi, Giovanbattista Gaspare; Gasparini, Sara; Cianci, Vittoria; Sueri, Chiara; Ferlazzo, Edoardo; Aguglia, Umberto

2017-03-01

A novel technique of quantitative EEG for differentiating patients with early-stage Creutzfeldt-Jakob disease (CJD) from other forms of rapidly progressive dementia (RPD) is proposed. The discrimination is based on the extraction of suitable features from the time-frequency representation of the EEG signals through continuous wavelet transform (CWT). An average measure of complexity of the EEG signal obtained by permutation entropy (PE) is also included. The dimensionality of the feature space is reduced through a multilayer processing system based on the recently emerged deep learning (DL) concept. The DL processor includes a stacked auto-encoder, trained by unsupervised learning techniques, and a classifier whose parameters are determined in a supervised way by associating the known category labels to the reduced vector of high-level features generated by the previous processing blocks. The supervised learning step is carried out by using either support vector machines (SVM) or multilayer neural networks (MLP-NN). A subset of EEG from patients suffering from Alzheimer's Disease (AD) and healthy controls (HC) is considered for differentiating CJD patients. When fine-tuning the parameters of the global processing system by a supervised learning procedure, the proposed system is able to achieve an average accuracy of 89%, an average sensitivity of 92%, and an average specificity of 89% in differentiating CJD from RPD. Similar results are obtained for CJD versus AD and CJD versus HC.

Incorporating High-Frequency Physiologic Data Using Computational Dictionary Learning Improves Prediction of Delayed Cerebral Ischemia Compared to Existing Methods.

PubMed

Megjhani, Murad; Terilli, Kalijah; Frey, Hans-Peter; Velazquez, Angela G; Doyle, Kevin William; Connolly, Edward Sander; Roh, David Jinou; Agarwal, Sachin; Claassen, Jan; Elhadad, Noemie; Park, Soojin

2018-01-01

Accurate prediction of delayed cerebral ischemia (DCI) after subarachnoid hemorrhage (SAH) can be critical for planning interventions to prevent poor neurological outcome. This paper presents a model using convolution dictionary learning to extract features from physiological data available from bedside monitors. We develop and validate a prediction model for DCI after SAH, demonstrating improved precision over standard methods alone. 488 consecutive SAH admissions from 2006 to 2014 to a tertiary care hospital were included. Models were trained on 80%, while 20% were set aside for validation testing. Modified Fisher Scale was considered the standard grading scale in clinical use; baseline features also analyzed included age, sex, Hunt-Hess, and Glasgow Coma Scales. An unsupervised approach using convolution dictionary learning was used to extract features from physiological time series (systolic blood pressure and diastolic blood pressure, heart rate, respiratory rate, and oxygen saturation). Classifiers (partial least squares and linear and kernel support vector machines) were trained on feature subsets of the derivation dataset. Models were applied to the validation dataset. The performances of the best classifiers on the validation dataset are reported by feature subset. Standard grading scale (mFS): AUC 0.54. Combined demographics and grading scales (baseline features): AUC 0.63. Kernel derived physiologic features: AUC 0.66. Combined baseline and physiologic features with redundant feature reduction: AUC 0.71 on derivation dataset and 0.78 on validation dataset. Current DCI prediction tools rely on admission imaging and are advantageously simple to employ. However, using an agnostic and computationally inexpensive learning approach for high-frequency physiologic time series data, we demonstrated that we could incorporate individual physiologic data to achieve higher classification accuracy.

Deep Unfolding for Topic Models.

PubMed

Chien, Jen-Tzung; Lee, Chao-Hsi

2018-02-01

Deep unfolding provides an approach to integrate the probabilistic generative models and the deterministic neural networks. Such an approach is benefited by deep representation, easy interpretation, flexible learning and stochastic modeling. This study develops the unsupervised and supervised learning of deep unfolded topic models for document representation and classification. Conventionally, the unsupervised and supervised topic models are inferred via the variational inference algorithm where the model parameters are estimated by maximizing the lower bound of logarithm of marginal likelihood using input documents without and with class labels, respectively. The representation capability or classification accuracy is constrained by the variational lower bound and the tied model parameters across inference procedure. This paper aims to relax these constraints by directly maximizing the end performance criterion and continuously untying the parameters in learning process via deep unfolding inference (DUI). The inference procedure is treated as the layer-wise learning in a deep neural network. The end performance is iteratively improved by using the estimated topic parameters according to the exponentiated updates. Deep learning of topic models is therefore implemented through a back-propagation procedure. Experimental results show the merits of DUI with increasing number of layers compared with variational inference in unsupervised as well as supervised topic models.

A proto-architecture for innate directionally selective visual maps.

PubMed

Adams, Samantha V; Harris, Chris M

2014-01-01

Self-organizing artificial neural networks are a popular tool for studying visual system development, in particular the cortical feature maps present in real systems that represent properties such as ocular dominance (OD), orientation-selectivity (OR) and direction selectivity (DS). They are also potentially useful in artificial systems, for example robotics, where the ability to extract and learn features from the environment in an unsupervised way is important. In this computational study we explore a DS map that is already latent in a simple artificial network. This latent selectivity arises purely from the cortical architecture without any explicit coding for DS and prior to any self-organising process facilitated by spontaneous activity or training. We find DS maps with local patchy regions that exhibit features similar to maps derived experimentally and from previous modeling studies. We explore the consequences of changes to the afferent and lateral connectivity to establish the key features of this proto-architecture that support DS.

Image quality classification for DR screening using deep learning.

PubMed

FengLi Yu; Jing Sun; Annan Li; Jun Cheng; Cheng Wan; Jiang Liu

2017-07-01

The quality of input images significantly affects the outcome of automated diabetic retinopathy (DR) screening systems. Unlike the previous methods that only consider simple low-level features such as hand-crafted geometric and structural features, in this paper we propose a novel method for retinal image quality classification (IQC) that performs computational algorithms imitating the working of the human visual system. The proposed algorithm combines unsupervised features from saliency map and supervised features coming from convolutional neural networks (CNN), which are fed to an SVM to automatically detect high quality vs poor quality retinal fundus images. We demonstrate the superior performance of our proposed algorithm on a large retinal fundus image dataset and the method could achieve higher accuracy than other methods. Although retinal images are used in this study, the methodology is applicable to the image quality assessment and enhancement of other types of medical images.

Massively parallel unsupervised single-particle cryo-EM data clustering via statistical manifold learning

PubMed Central

Wu, Jiayi; Ma, Yong-Bei; Congdon, Charles; Brett, Bevin; Chen, Shuobing; Xu, Yaofang; Ouyang, Qi

2017-01-01

Structural heterogeneity in single-particle cryo-electron microscopy (cryo-EM) data represents a major challenge for high-resolution structure determination. Unsupervised classification may serve as the first step in the assessment of structural heterogeneity. However, traditional algorithms for unsupervised classification, such as K-means clustering and maximum likelihood optimization, may classify images into wrong classes with decreasing signal-to-noise-ratio (SNR) in the image data, yet demand increased computational costs. Overcoming these limitations requires further development of clustering algorithms for high-performance cryo-EM data processing. Here we introduce an unsupervised single-particle clustering algorithm derived from a statistical manifold learning framework called generative topographic mapping (GTM). We show that unsupervised GTM clustering improves classification accuracy by about 40% in the absence of input references for data with lower SNRs. Applications to several experimental datasets suggest that our algorithm can detect subtle structural differences among classes via a hierarchical clustering strategy. After code optimization over a high-performance computing (HPC) environment, our software implementation was able to generate thousands of reference-free class averages within hours in a massively parallel fashion, which allows a significant improvement on ab initio 3D reconstruction and assists in the computational purification of homogeneous datasets for high-resolution visualization. PMID:28786986

Massively parallel unsupervised single-particle cryo-EM data clustering via statistical manifold learning.

PubMed

Wu, Jiayi; Ma, Yong-Bei; Congdon, Charles; Brett, Bevin; Chen, Shuobing; Xu, Yaofang; Ouyang, Qi; Mao, Youdong

2017-01-01

Structural heterogeneity in single-particle cryo-electron microscopy (cryo-EM) data represents a major challenge for high-resolution structure determination. Unsupervised classification may serve as the first step in the assessment of structural heterogeneity. However, traditional algorithms for unsupervised classification, such as K-means clustering and maximum likelihood optimization, may classify images into wrong classes with decreasing signal-to-noise-ratio (SNR) in the image data, yet demand increased computational costs. Overcoming these limitations requires further development of clustering algorithms for high-performance cryo-EM data processing. Here we introduce an unsupervised single-particle clustering algorithm derived from a statistical manifold learning framework called generative topographic mapping (GTM). We show that unsupervised GTM clustering improves classification accuracy by about 40% in the absence of input references for data with lower SNRs. Applications to several experimental datasets suggest that our algorithm can detect subtle structural differences among classes via a hierarchical clustering strategy. After code optimization over a high-performance computing (HPC) environment, our software implementation was able to generate thousands of reference-free class averages within hours in a massively parallel fashion, which allows a significant improvement on ab initio 3D reconstruction and assists in the computational purification of homogeneous datasets for high-resolution visualization.

Deep learning-based subdivision approach for large scale macromolecules structure recovery from electron cryo tomograms

PubMed Central

Xu, Min; Chai, Xiaoqi; Muthakana, Hariank; Liang, Xiaodan; Yang, Ge; Zeev-Ben-Mordehai, Tzviya; Xing, Eric P.

2017-01-01

Abstract Motivation: Cellular Electron CryoTomography (CECT) enables 3D visualization of cellular organization at near-native state and in sub-molecular resolution, making it a powerful tool for analyzing structures of macromolecular complexes and their spatial organizations inside single cells. However, high degree of structural complexity together with practical imaging limitations makes the systematic de novo discovery of structures within cells challenging. It would likely require averaging and classifying millions of subtomograms potentially containing hundreds of highly heterogeneous structural classes. Although it is no longer difficult to acquire CECT data containing such amount of subtomograms due to advances in data acquisition automation, existing computational approaches have very limited scalability or discrimination ability, making them incapable of processing such amount of data. Results: To complement existing approaches, in this article we propose a new approach for subdividing subtomograms into smaller but relatively homogeneous subsets. The structures in these subsets can then be separately recovered using existing computation intensive methods. Our approach is based on supervised structural feature extraction using deep learning, in combination with unsupervised clustering and reference-free classification. Our experiments show that, compared with existing unsupervised rotation invariant feature and pose-normalization based approaches, our new approach achieves significant improvements in both discrimination ability and scalability. More importantly, our new approach is able to discover new structural classes and recover structures that do not exist in training data. Availability and Implementation: Source code freely available at http://www.cs.cmu.edu/∼mxu1/software. Contact: mxu1@cs.cmu.edu Supplementary information: Supplementary data are available at Bioinformatics online. PMID:28881965

Unsupervised learning of facial emotion decoding skills.

PubMed

Huelle, Jan O; Sack, Benjamin; Broer, Katja; Komlewa, Irina; Anders, Silke

2014-01-01

Research on the mechanisms underlying human facial emotion recognition has long focussed on genetically determined neural algorithms and often neglected the question of how these algorithms might be tuned by social learning. Here we show that facial emotion decoding skills can be significantly and sustainably improved by practice without an external teaching signal. Participants saw video clips of dynamic facial expressions of five different women and were asked to decide which of four possible emotions (anger, disgust, fear, and sadness) was shown in each clip. Although no external information about the correctness of the participant's response or the sender's true affective state was provided, participants showed a significant increase of facial emotion recognition accuracy both within and across two training sessions two days to several weeks apart. We discuss several similarities and differences between the unsupervised improvement of facial decoding skills observed in the current study, unsupervised perceptual learning of simple stimuli described in previous studies and practice effects often observed in cognitive tasks.

Unsupervised learning of facial emotion decoding skills

PubMed Central

Huelle, Jan O.; Sack, Benjamin; Broer, Katja; Komlewa, Irina; Anders, Silke

2013-01-01

Research on the mechanisms underlying human facial emotion recognition has long focussed on genetically determined neural algorithms and often neglected the question of how these algorithms might be tuned by social learning. Here we show that facial emotion decoding skills can be significantly and sustainably improved by practice without an external teaching signal. Participants saw video clips of dynamic facial expressions of five different women and were asked to decide which of four possible emotions (anger, disgust, fear, and sadness) was shown in each clip. Although no external information about the correctness of the participant’s response or the sender’s true affective state was provided, participants showed a significant increase of facial emotion recognition accuracy both within and across two training sessions two days to several weeks apart. We discuss several similarities and differences between the unsupervised improvement of facial decoding skills observed in the current study, unsupervised perceptual learning of simple visual stimuli described in previous studies and practice effects often observed in cognitive tasks. PMID:24578686

Self-organizing neural networks--an alternative way of cluster analysis in clinical chemistry.

PubMed

Reibnegger, G; Wachter, H

1996-04-15

Supervised learning schemes have been employed by several workers for training neural networks designed to solve clinical problems. We demonstrate that unsupervised techniques can also produce interesting and meaningful results. Using a data set on the chemical composition of milk from 22 different mammals, we demonstrate that self-organizing feature maps (Kohonen networks) as well as a modified version of error backpropagation technique yield results mimicking conventional cluster analysis. Both techniques are able to project a potentially multi-dimensional input vector onto a two-dimensional space whereby neighborhood relationships remain conserved. Thus, these techniques can be used for reducing dimensionality of complicated data sets and for enhancing comprehensibility of features hidden in the data matrix.

PANTHER. Trajectory Analysis

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

Rintoul, Mark Daniel; Wilson, Andrew T.; Valicka, Christopher G.

We want to organize a body of trajectories in order to identify, search for, classify and predict behavior among objects such as aircraft and ships. Existing compari- son functions such as the Fr'echet distance are computationally expensive and yield counterintuitive results in some cases. We propose an approach using feature vectors whose components represent succinctly the salient information in trajectories. These features incorporate basic information such as total distance traveled and distance be- tween start/stop points as well as geometric features related to the properties of the convex hull, trajectory curvature and general distance geometry. Additionally, these features can generallymore » be mapped easily to behaviors of interest to humans that are searching large databases. Most of these geometric features are invariant under rigid transformation. We demonstrate the use of different subsets of these features to iden- tify trajectories similar to an exemplar, cluster a database of several hundred thousand trajectories, predict destination and apply unsupervised machine learning algorithms.« less

Bilingual Lexical Interactions in an Unsupervised Neural Network Model

ERIC Educational Resources Information Center

Zhao, Xiaowei; Li, Ping

2010-01-01

In this paper we present an unsupervised neural network model of bilingual lexical development and interaction. We focus on how the representational structures of the bilingual lexicons can emerge, develop, and interact with each other as a function of the learning history. The results show that: (1) distinct representations for the two lexicons…

An illustration of new methods in machine condition monitoring, Part I: stochastic resonance

NASA Astrophysics Data System (ADS)

Worden, K.; Antoniadou, I.; Marchesiello, S.; Mba, C.; Garibaldi, L.

2017-05-01

There have been many recent developments in the application of data-based methods to machine condition monitoring. A powerful methodology based on machine learning has emerged, where diagnostics are based on a two-step procedure: extraction of damage-sensitive features, followed by unsupervised learning (novelty detection) or supervised learning (classification). The objective of the current pair of papers is simply to illustrate one state-of-the-art procedure for each step, using synthetic data representative of reality in terms of size and complexity. The first paper in the pair will deal with feature extraction. Although some papers have appeared in the recent past considering stochastic resonance as a means of amplifying damage information in signals, they have largely relied on ad hoc specifications of the resonator used. In contrast, the current paper will adopt a principled optimisation-based approach to the resonator design. The paper will also show that a discrete dynamical system can provide all the benefits of a continuous system, but also provide a considerable speed-up in terms of simulation time in order to facilitate the optimisation approach.

Unsupervised learning of natural languages

PubMed Central

Solan, Zach; Horn, David; Ruppin, Eytan; Edelman, Shimon

2005-01-01

We address the problem, fundamental to linguistics, bioinformatics, and certain other disciplines, of using corpora of raw symbolic sequential data to infer underlying rules that govern their production. Given a corpus of strings (such as text, transcribed speech, chromosome or protein sequence data, sheet music, etc.), our unsupervised algorithm recursively distills from it hierarchically structured patterns. The adios (automatic distillation of structure) algorithm relies on a statistical method for pattern extraction and on structured generalization, two processes that have been implicated in language acquisition. It has been evaluated on artificial context-free grammars with thousands of rules, on natural languages as diverse as English and Chinese, and on protein data correlating sequence with function. This unsupervised algorithm is capable of learning complex syntax, generating grammatical novel sentences, and proving useful in other fields that call for structure discovery from raw data, such as bioinformatics. PMID:16087885

Unsupervised learning of natural languages.

PubMed

Solan, Zach; Horn, David; Ruppin, Eytan; Edelman, Shimon

2005-08-16

We address the problem, fundamental to linguistics, bioinformatics, and certain other disciplines, of using corpora of raw symbolic sequential data to infer underlying rules that govern their production. Given a corpus of strings (such as text, transcribed speech, chromosome or protein sequence data, sheet music, etc.), our unsupervised algorithm recursively distills from it hierarchically structured patterns. The adios (automatic distillation of structure) algorithm relies on a statistical method for pattern extraction and on structured generalization, two processes that have been implicated in language acquisition. It has been evaluated on artificial context-free grammars with thousands of rules, on natural languages as diverse as English and Chinese, and on protein data correlating sequence with function. This unsupervised algorithm is capable of learning complex syntax, generating grammatical novel sentences, and proving useful in other fields that call for structure discovery from raw data, such as bioinformatics.

Classification and unsupervised clustering of LIGO data with Deep Transfer Learning

NASA Astrophysics Data System (ADS)

George, Daniel; Shen, Hongyu; Huerta, E. A.

2018-05-01

Gravitational wave detection requires a detailed understanding of the response of the LIGO and Virgo detectors to true signals in the presence of environmental and instrumental noise. Of particular interest is the study of anomalous non-Gaussian transients, such as glitches, since their occurrence rate in LIGO and Virgo data can obscure or even mimic true gravitational wave signals. Therefore, successfully identifying and excising these anomalies from gravitational wave data is of utmost importance for the detection and characterization of true signals and for the accurate computation of their significance. To facilitate this work, we present the first application of deep learning combined with transfer learning to show that knowledge from pretrained models for real-world object recognition can be transferred for classifying spectrograms of glitches. To showcase this new method, we use a data set of twenty-two classes of glitches, curated and labeled by the Gravity Spy project using data collected during LIGO's first discovery campaign. We demonstrate that our Deep Transfer Learning method enables an optimal use of very deep convolutional neural networks for glitch classification given small and unbalanced training data sets, significantly reduces the training time, and achieves state-of-the-art accuracy above 98.8%, lowering the previous error rate by over 60%. More importantly, once trained via transfer learning on the known classes, we show that our neural networks can be truncated and used as feature extractors for unsupervised clustering to automatically group together new unknown classes of glitches and anomalous signals. This novel capability is of paramount importance to identify and remove new types of glitches which will occur as the LIGO/Virgo detectors gradually attain design sensitivity.

Unsupervised Clustering of Subcellular Protein Expression Patterns in High-Throughput Microscopy Images Reveals Protein Complexes and Functional Relationships between Proteins

PubMed Central

Handfield, Louis-François; Chong, Yolanda T.; Simmons, Jibril; Andrews, Brenda J.; Moses, Alan M.

2013-01-01

Protein subcellular localization has been systematically characterized in budding yeast using fluorescently tagged proteins. Based on the fluorescence microscopy images, subcellular localization of many proteins can be classified automatically using supervised machine learning approaches that have been trained to recognize predefined image classes based on statistical features. Here, we present an unsupervised analysis of protein expression patterns in a set of high-resolution, high-throughput microscope images. Our analysis is based on 7 biologically interpretable features which are evaluated on automatically identified cells, and whose cell-stage dependency is captured by a continuous model for cell growth. We show that it is possible to identify most previously identified localization patterns in a cluster analysis based on these features and that similarities between the inferred expression patterns contain more information about protein function than can be explained by a previous manual categorization of subcellular localization. Furthermore, the inferred cell-stage associated to each fluorescence measurement allows us to visualize large groups of proteins entering the bud at specific stages of bud growth. These correspond to proteins localized to organelles, revealing that the organelles must be entering the bud in a stereotypical order. We also identify and organize a smaller group of proteins that show subtle differences in the way they move around the bud during growth. Our results suggest that biologically interpretable features based on explicit models of cell morphology will yield unprecedented power for pattern discovery in high-resolution, high-throughput microscopy images. PMID:23785265

A Learning Model for L/M Specificity in Ganglion Cells

NASA Technical Reports Server (NTRS)

Ahumada, Albert J.

2016-01-01

An unsupervised learning model for developing LM specific wiring at the ganglion cell level would support the research indicating LM specific wiring at the ganglion cell level (Reid and Shapley, 2002). Removing the contributions to the surround from cells of the same cone type improves the signal-to-noise ratio of the chromatic signals. The unsupervised learning model used is Hebbian associative learning, which strengthens the surround input connections according to the correlation of the output with the input. Since the surround units of the same cone type as the center are redundant with the center, their weights end up disappearing. This process can be thought of as a general mechanism for eliminating unnecessary cells in the nervous system.

Epithelium-Stroma Classification via Convolutional Neural Networks and Unsupervised Domain Adaptation in Histopathological Images.

PubMed

Huang, Yue; Zheng, Han; Liu, Chi; Ding, Xinghao; Rohde, Gustavo K

2017-11-01

Epithelium-stroma classification is a necessary preprocessing step in histopathological image analysis. Current deep learning based recognition methods for histology data require collection of large volumes of labeled data in order to train a new neural network when there are changes to the image acquisition procedure. However, it is extremely expensive for pathologists to manually label sufficient volumes of data for each pathology study in a professional manner, which results in limitations in real-world applications. A very simple but effective deep learning method, that introduces the concept of unsupervised domain adaptation to a simple convolutional neural network (CNN), has been proposed in this paper. Inspired by transfer learning, our paper assumes that the training data and testing data follow different distributions, and there is an adaptation operation to more accurately estimate the kernels in CNN in feature extraction, in order to enhance performance by transferring knowledge from labeled data in source domain to unlabeled data in target domain. The model has been evaluated using three independent public epithelium-stroma datasets by cross-dataset validations. The experimental results demonstrate that for epithelium-stroma classification, the proposed framework outperforms the state-of-the-art deep neural network model, and it also achieves better performance than other existing deep domain adaptation methods. The proposed model can be considered to be a better option for real-world applications in histopathological image analysis, since there is no longer a requirement for large-scale labeled data in each specified domain.

Unsupervised chunking based on graph propagation from bilingual corpus.

PubMed

Zhu, Ling; Wong, Derek F; Chao, Lidia S

2014-01-01

This paper presents a novel approach for unsupervised shallow parsing model trained on the unannotated Chinese text of parallel Chinese-English corpus. In this approach, no information of the Chinese side is applied. The exploitation of graph-based label propagation for bilingual knowledge transfer, along with an application of using the projected labels as features in unsupervised model, contributes to a better performance. The experimental comparisons with the state-of-the-art algorithms show that the proposed approach is able to achieve impressive higher accuracy in terms of F-score.

Latent feature representation with stacked auto-encoder for AD/MCI diagnosis

PubMed Central

Lee, Seong-Whan

2014-01-01

Recently, there have been great interests for computer-aided diagnosis of Alzheimer’s disease (AD) and its prodromal stage, mild cognitive impairment (MCI). Unlike the previous methods that considered simple low-level features such as gray matter tissue volumes from MRI, and mean signal intensities from PET, in this paper, we propose a deep learning-based latent feature representation with a stacked auto-encoder (SAE). We believe that there exist latent non-linear complicated patterns inherent in the low-level features such as relations among features. Combining the latent information with the original features helps build a robust model in AD/MCI classification, with high diagnostic accuracy. Furthermore, thanks to the unsupervised characteristic of the pre-training in deep learning, we can benefit from the target-unrelated samples to initialize parameters of SAE, thus finding optimal parameters in fine-tuning with the target-related samples, and further enhancing the classification performances across four binary classification problems: AD vs. healthy normal control (HC), MCI vs. HC, AD vs. MCI, and MCI converter (MCI-C) vs. MCI non-converter (MCI-NC). In our experiments on ADNI dataset, we validated the effectiveness of the proposed method, showing the accuracies of 98.8, 90.7, 83.7, and 83.3 % for AD/HC, MCI/HC, AD/MCI, and MCI-C/MCI-NC classification, respectively. We believe that deep learning can shed new light on the neuroimaging data analysis, and our work presented the applicability of this method to brain disease diagnosis. PMID:24363140

Extraction of features from medical images using a modular neural network approach that relies on learning by sample

NASA Astrophysics Data System (ADS)

Brahmi, Djamel; Serruys, Camille; Cassoux, Nathalie; Giron, Alain; Triller, Raoul; Lehoang, Phuc; Fertil, Bernard

2000-06-01

Medical images provide experienced physicians with meaningful visual stimuli but their features are frequently hard to decipher. The development of a computational model to mimic physicians' expertise is a demanding task, especially if a significant and sophisticated preprocessing of images is required. Learning from well-expertised images may be a more convenient approach, inasmuch a large and representative bunch of samples is available. A four-stage approach has been designed, which combines image sub-sampling with unsupervised image coding, supervised classification and image reconstruction in order to directly extract medical expertise from raw images. The system has been applied (1) to the detection of some features related to the diagnosis of black tumors of skin (a classification issue) and (2) to the detection of virus-infected and healthy areas in retina angiography in order to locate precisely the border between them and characterize the evolution of infection. For reasonably balanced training sets, we are able to obtained about 90% correct classification of features (black tumors). Boundaries generated by our system mimic reproducibility of hand-outlines drawn by experts (segmentation of virus-infected area).

Unsupervised laparoscopic appendicectomy by surgical trainees is safe and time-effective.

PubMed

Wong, Kenneth; Duncan, Tristram; Pearson, Andrew

2007-07-01

Open appendicectomy is the traditional standard treatment for appendicitis. Laparoscopic appendicectomy is perceived as a procedure with greater potential for complications and longer operative times. This paper examines the hypothesis that unsupervised laparoscopic appendicectomy by surgical trainees is a safe and time-effective valid alternative. Medical records, operating theatre records and histopathology reports of all patients undergoing laparoscopic and open appendicectomy over a 15-month period in two hospitals within an area health service were retrospectively reviewed. Data were analysed to compare patient features, pathology findings, operative times, complications, readmissions and mortality between laparoscopic and open groups and between unsupervised surgical trainee operators versus consultant surgeon operators. A total of 143 laparoscopic and 222 open appendicectomies were reviewed. Unsupervised trainees performed 64% of the laparoscopic appendicectomies and 55% of the open appendicectomies. There were no significant differences in complication rates, readmissions, mortality and length of stay between laparoscopic and open appendicectomy groups or between trainee and consultant surgeon operators. Conversion rates (laparoscopic to open approach) were similar for trainees and consultants. Unsupervised senior surgical trainees did not take significantly longer to perform laparoscopic appendicectomy when compared to unsupervised trainee-performed open appendicectomy. Unsupervised laparoscopic appendicectomy by surgical trainees is safe and time-effective.

Interactive classification and content-based retrieval of tissue images

NASA Astrophysics Data System (ADS)

Aksoy, Selim; Marchisio, Giovanni B.; Tusk, Carsten; Koperski, Krzysztof

2002-11-01

We describe a system for interactive classification and retrieval of microscopic tissue images. Our system models tissues in pixel, region and image levels. Pixel level features are generated using unsupervised clustering of color and texture values. Region level features include shape information and statistics of pixel level feature values. Image level features include statistics and spatial relationships of regions. To reduce the gap between low-level features and high-level expert knowledge, we define the concept of prototype regions. The system learns the prototype regions in an image collection using model-based clustering and density estimation. Different tissue types are modeled using spatial relationships of these regions. Spatial relationships are represented by fuzzy membership functions. The system automatically selects significant relationships from training data and builds models which can also be updated using user relevance feedback. A Bayesian framework is used to classify tissues based on these models. Preliminary experiments show that the spatial relationship models we developed provide a flexible and powerful framework for classification and retrieval of tissue images.

Manifold Learning in MR spectroscopy using nonlinear dimensionality reduction and unsupervised clustering.

PubMed

Yang, Guang; Raschke, Felix; Barrick, Thomas R; Howe, Franklyn A

2015-09-01

To investigate whether nonlinear dimensionality reduction improves unsupervised classification of (1) H MRS brain tumor data compared with a linear method. In vivo single-voxel (1) H magnetic resonance spectroscopy (55 patients) and (1) H magnetic resonance spectroscopy imaging (MRSI) (29 patients) data were acquired from histopathologically diagnosed gliomas. Data reduction using Laplacian eigenmaps (LE) or independent component analysis (ICA) was followed by k-means clustering or agglomerative hierarchical clustering (AHC) for unsupervised learning to assess tumor grade and for tissue type segmentation of MRSI data. An accuracy of 93% in classification of glioma grade II and grade IV, with 100% accuracy in distinguishing tumor and normal spectra, was obtained by LE with unsupervised clustering, but not with the combination of k-means and ICA. With (1) H MRSI data, LE provided a more linear distribution of data for cluster analysis and better cluster stability than ICA. LE combined with k-means or AHC provided 91% accuracy for classifying tumor grade and 100% accuracy for identifying normal tissue voxels. Color-coded visualization of normal brain, tumor core, and infiltration regions was achieved with LE combined with AHC. The LE method is promising for unsupervised clustering to separate brain and tumor tissue with automated color-coding for visualization of (1) H MRSI data after cluster analysis. © 2014 Wiley Periodicals, Inc.

Scaling up spike-and-slab models for unsupervised feature learning.

PubMed

Goodfellow, Ian J; Courville, Aaron; Bengio, Yoshua

2013-08-01

We describe the use of two spike-and-slab models for modeling real-valued data, with an emphasis on their applications to object recognition. The first model, which we call spike-and-slab sparse coding (S3C), is a preexisting model for which we introduce a faster approximate inference algorithm. We introduce a deep variant of S3C, which we call the partially directed deep Boltzmann machine (PD-DBM) and extend our S3C inference algorithm for use on this model. We describe learning procedures for each. We demonstrate that our inference procedure for S3C enables scaling the model to unprecedented large problem sizes, and demonstrate that using S3C as a feature extractor results in very good object recognition performance, particularly when the number of labeled examples is low. We show that the PD-DBM generates better samples than its shallow counterpart, and that unlike DBMs or DBNs, the PD-DBM may be trained successfully without greedy layerwise training.

Comprehensive machine learning analysis of Hydra behavior reveals a stable basal behavioral repertoire.

PubMed

Han, Shuting; Taralova, Ekaterina; Dupre, Christophe; Yuste, Rafael

2018-03-28

Animal behavior has been studied for centuries, but few efficient methods are available to automatically identify and classify it. Quantitative behavioral studies have been hindered by the subjective and imprecise nature of human observation, and the slow speed of annotating behavioral data. Here, we developed an automatic behavior analysis pipeline for the cnidarian Hydra vulgaris using machine learning. We imaged freely behaving Hydra , extracted motion and shape features from the videos, and constructed a dictionary of visual features to classify pre-defined behaviors. We also identified unannotated behaviors with unsupervised methods. Using this analysis pipeline, we quantified 6 basic behaviors and found surprisingly similar behavior statistics across animals within the same species, regardless of experimental conditions. Our analysis indicates that the fundamental behavioral repertoire of Hydra is stable. This robustness could reflect a homeostatic neural control of "housekeeping" behaviors which could have been already present in the earliest nervous systems. © 2018, Han et al.

Feature Selection for Ridge Regression with Provable Guarantees.

PubMed

Paul, Saurabh; Drineas, Petros

2016-04-01

We introduce single-set spectral sparsification as a deterministic sampling-based feature selection technique for regularized least-squares classification, which is the classification analog to ridge regression. The method is unsupervised and gives worst-case guarantees of the generalization power of the classification function after feature selection with respect to the classification function obtained using all features. We also introduce leverage-score sampling as an unsupervised randomized feature selection method for ridge regression. We provide risk bounds for both single-set spectral sparsification and leverage-score sampling on ridge regression in the fixed design setting and show that the risk in the sampled space is comparable to the risk in the full-feature space. We perform experiments on synthetic and real-world data sets; a subset of TechTC-300 data sets, to support our theory. Experimental results indicate that the proposed methods perform better than the existing feature selection methods.

Pharmacovigilance from social media: mining adverse drug reaction mentions using sequence labeling with word embedding cluster features.

PubMed

Nikfarjam, Azadeh; Sarker, Abeed; O'Connor, Karen; Ginn, Rachel; Gonzalez, Graciela

2015-05-01

Social media is becoming increasingly popular as a platform for sharing personal health-related information. This information can be utilized for public health monitoring tasks, particularly for pharmacovigilance, via the use of natural language processing (NLP) techniques. However, the language in social media is highly informal, and user-expressed medical concepts are often nontechnical, descriptive, and challenging to extract. There has been limited progress in addressing these challenges, and thus far, advanced machine learning-based NLP techniques have been underutilized. Our objective is to design a machine learning-based approach to extract mentions of adverse drug reactions (ADRs) from highly informal text in social media. We introduce ADRMine, a machine learning-based concept extraction system that uses conditional random fields (CRFs). ADRMine utilizes a variety of features, including a novel feature for modeling words' semantic similarities. The similarities are modeled by clustering words based on unsupervised, pretrained word representation vectors (embeddings) generated from unlabeled user posts in social media using a deep learning technique. ADRMine outperforms several strong baseline systems in the ADR extraction task by achieving an F-measure of 0.82. Feature analysis demonstrates that the proposed word cluster features significantly improve extraction performance. It is possible to extract complex medical concepts, with relatively high performance, from informal, user-generated content. Our approach is particularly scalable, suitable for social media mining, as it relies on large volumes of unlabeled data, thus diminishing the need for large, annotated training data sets. © The Author 2015. Published by Oxford University Press on behalf of the American Medical Informatics Association.

Ellipsoidal fuzzy learning for smart car platoons

NASA Astrophysics Data System (ADS)

Dickerson, Julie A.; Kosko, Bart

1993-12-01

A neural-fuzzy system combined supervised and unsupervised learning to find and tune the fuzzy-rules. An additive fuzzy system approximates a function by covering its graph with fuzzy rules. A fuzzy rule patch can take the form of an ellipsoid in the input-output space. Unsupervised competitive learning found the statistics of data clusters. The covariance matrix of each synaptic quantization vector defined on ellipsoid centered at the centroid of the data cluster. Tightly clustered data gave smaller ellipsoids or more certain rules. Sparse data gave larger ellipsoids or less certain rules. Supervised learning tuned the ellipsoids to improve the approximation. The supervised neural system used gradient descent to find the ellipsoidal fuzzy patches. It locally minimized the mean-squared error of the fuzzy approximation. Hybrid ellipsoidal learning estimated the control surface for a smart car controller.

The New Possibilities from "Big Data" to Overlooked Associations Between Diabetes, Biochemical Parameters, Glucose Control, and Osteoporosis.

PubMed

Kruse, Christian

2018-06-01

To review current practices and technologies within the scope of "Big Data" that can further our understanding of diabetes mellitus and osteoporosis from large volumes of data. "Big Data" techniques involving supervised machine learning, unsupervised machine learning, and deep learning image analysis are presented with examples of current literature. Supervised machine learning can allow us to better predict diabetes-induced osteoporosis and understand relative predictor importance of diabetes-affected bone tissue. Unsupervised machine learning can allow us to understand patterns in data between diabetic pathophysiology and altered bone metabolism. Image analysis using deep learning can allow us to be less dependent on surrogate predictors and use large volumes of images to classify diabetes-induced osteoporosis and predict future outcomes directly from images. "Big Data" techniques herald new possibilities to understand diabetes-induced osteoporosis and ascertain our current ability to classify, understand, and predict this condition.

Neural network-based multiple robot simultaneous localization and mapping.

PubMed

Saeedi, Sajad; Paull, Liam; Trentini, Michael; Li, Howard

2011-12-01

In this paper, a decentralized platform for simultaneous localization and mapping (SLAM) with multiple robots is developed. Each robot performs single robot view-based SLAM using an extended Kalman filter to fuse data from two encoders and a laser ranger. To extend this approach to multiple robot SLAM, a novel occupancy grid map fusion algorithm is proposed. Map fusion is achieved through a multistep process that includes image preprocessing, map learning (clustering) using neural networks, relative orientation extraction using norm histogram cross correlation and a Radon transform, relative translation extraction using matching norm vectors, and then verification of the results. The proposed map learning method is a process based on the self-organizing map. In the learning phase, the obstacles of the map are learned by clustering the occupied cells of the map into clusters. The learning is an unsupervised process which can be done on the fly without any need to have output training patterns. The clusters represent the spatial form of the map and make further analyses of the map easier and faster. Also, clusters can be interpreted as features extracted from the occupancy grid map so the map fusion problem becomes a task of matching features. Results of the experiments from tests performed on a real environment with multiple robots prove the effectiveness of the proposed solution.

Efficient coding of spectrotemporal binaural sounds leads to emergence of the auditory space representation

PubMed Central

Młynarski, Wiktor

2014-01-01

To date a number of studies have shown that receptive field shapes of early sensory neurons can be reproduced by optimizing coding efficiency of natural stimulus ensembles. A still unresolved question is whether the efficient coding hypothesis explains formation of neurons which explicitly represent environmental features of different functional importance. This paper proposes that the spatial selectivity of higher auditory neurons emerges as a direct consequence of learning efficient codes for natural binaural sounds. Firstly, it is demonstrated that a linear efficient coding transform—Independent Component Analysis (ICA) trained on spectrograms of naturalistic simulated binaural sounds extracts spatial information present in the signal. A simple hierarchical ICA extension allowing for decoding of sound position is proposed. Furthermore, it is shown that units revealing spatial selectivity can be learned from a binaural recording of a natural auditory scene. In both cases a relatively small subpopulation of learned spectrogram features suffices to perform accurate sound localization. Representation of the auditory space is therefore learned in a purely unsupervised way by maximizing the coding efficiency and without any task-specific constraints. This results imply that efficient coding is a useful strategy for learning structures which allow for making behaviorally vital inferences about the environment. PMID:24639644

Unsupervised quality estimation model for English to German translation and its application in extensive supervised evaluation.

PubMed

Han, Aaron L-F; Wong, Derek F; Chao, Lidia S; He, Liangye; Lu, Yi

2014-01-01

With the rapid development of machine translation (MT), the MT evaluation becomes very important to timely tell us whether the MT system makes any progress. The conventional MT evaluation methods tend to calculate the similarity between hypothesis translations offered by automatic translation systems and reference translations offered by professional translators. There are several weaknesses in existing evaluation metrics. Firstly, the designed incomprehensive factors result in language-bias problem, which means they perform well on some special language pairs but weak on other language pairs. Secondly, they tend to use no linguistic features or too many linguistic features, of which no usage of linguistic feature draws a lot of criticism from the linguists and too many linguistic features make the model weak in repeatability. Thirdly, the employed reference translations are very expensive and sometimes not available in the practice. In this paper, the authors propose an unsupervised MT evaluation metric using universal part-of-speech tagset without relying on reference translations. The authors also explore the performances of the designed metric on traditional supervised evaluation tasks. Both the supervised and unsupervised experiments show that the designed methods yield higher correlation scores with human judgments.

Wavelet-based unsupervised learning method for electrocardiogram suppression in surface electromyograms.

PubMed

Niegowski, Maciej; Zivanovic, Miroslav

2016-03-01

We present a novel approach aimed at removing electrocardiogram (ECG) perturbation from single-channel surface electromyogram (EMG) recordings by means of unsupervised learning of wavelet-based intensity images. The general idea is to combine the suitability of certain wavelet decomposition bases which provide sparse electrocardiogram time-frequency representations, with the capacity of non-negative matrix factorization (NMF) for extracting patterns from images. In order to overcome convergence problems which often arise in NMF-related applications, we design a novel robust initialization strategy which ensures proper signal decomposition in a wide range of ECG contamination levels. Moreover, the method can be readily used because no a priori knowledge or parameter adjustment is needed. The proposed method was evaluated on real surface EMG signals against two state-of-the-art unsupervised learning algorithms and a singular spectrum analysis based method. The results, expressed in terms of high-to-low energy ratio, normalized median frequency, spectral power difference and normalized average rectified value, suggest that the proposed method enables better ECG-EMG separation quality than the reference methods. Copyright © 2015 IPEM. Published by Elsevier Ltd. All rights reserved.

Galaxy morphology - An unsupervised machine learning approach

NASA Astrophysics Data System (ADS)

Schutter, A.; Shamir, L.

2015-09-01

Structural properties poses valuable information about the formation and evolution of galaxies, and are important for understanding the past, present, and future universe. Here we use unsupervised machine learning methodology to analyze a network of similarities between galaxy morphological types, and automatically deduce a morphological sequence of galaxies. Application of the method to the EFIGI catalog show that the morphological scheme produced by the algorithm is largely in agreement with the De Vaucouleurs system, demonstrating the ability of computer vision and machine learning methods to automatically profile galaxy morphological sequences. The unsupervised analysis method is based on comprehensive computer vision techniques that compute the visual similarities between the different morphological types. Rather than relying on human cognition, the proposed system deduces the similarities between sets of galaxy images in an automatic manner, and is therefore not limited by the number of galaxies being analyzed. The source code of the method is publicly available, and the protocol of the experiment is included in the paper so that the experiment can be replicated, and the method can be used to analyze user-defined datasets of galaxy images.

Unsupervised segmentation of lungs from chest radiographs

NASA Astrophysics Data System (ADS)

Ghosh, Payel; Antani, Sameer K.; Long, L. Rodney; Thoma, George R.

2012-03-01

This paper describes our preliminary investigations for deriving and characterizing coarse-level textural regions present in the lung field on chest radiographs using unsupervised grow-cut (UGC), a cellular automaton based unsupervised segmentation technique. The segmentation has been performed on a publicly available data set of chest radiographs. The algorithm is useful for this application because it automatically converges to a natural segmentation of the image from random seed points using low-level image features such as pixel intensity values and texture features. Our goal is to develop a portable screening system for early detection of lung diseases for use in remote areas in developing countries. This involves developing automated algorithms for screening x-rays as normal/abnormal with a high degree of sensitivity, and identifying lung disease patterns on chest x-rays. Automatically deriving and quantitatively characterizing abnormal regions present in the lung field is the first step toward this goal. Therefore, region-based features such as geometrical and pixel-value measurements were derived from the segmented lung fields. In the future, feature selection and classification will be performed to identify pathological conditions such as pulmonary tuberculosis on chest radiographs. Shape-based features will also be incorporated to account for occlusions of the lung field and by other anatomical structures such as the heart and diaphragm.

Taxonomy-aware feature engineering for microbiome classification.

PubMed

Oudah, Mai; Henschel, Andreas

2018-06-15

What is a healthy microbiome? The pursuit of this and many related questions, especially in light of the recently recognized microbial component in a wide range of diseases has sparked a surge in metagenomic studies. They are often not simply attributable to a single pathogen but rather are the result of complex ecological processes. Relatedly, the increasing DNA sequencing depth and number of samples in metagenomic case-control studies enabled the applicability of powerful statistical methods, e.g. Machine Learning approaches. For the latter, the feature space is typically shaped by the relative abundances of operational taxonomic units, as determined by cost-effective phylogenetic marker gene profiles. While a substantial body of microbiome/microbiota research involves unsupervised and supervised Machine Learning, very little attention has been put on feature selection and engineering. We here propose the first algorithm to exploit phylogenetic hierarchy (i.e. an all-encompassing taxonomy) in feature engineering for microbiota classification. The rationale is to exploit the often mono- or oligophyletic distribution of relevant (but hidden) traits by virtue of taxonomic abstraction. The algorithm is embedded in a comprehensive microbiota classification pipeline, which we applied to a diverse range of datasets, distinguishing healthy from diseased microbiota samples. We demonstrate substantial improvements over the state-of-the-art microbiota classification tools in terms of classification accuracy, regardless of the actual Machine Learning technique while using drastically reduced feature spaces. Moreover, generalized features bear great explanatory value: they provide a concise description of conditions and thus help to provide pathophysiological insights. Indeed, the automatically and reproducibly derived features are consistent with previously published domain expert analyses.

Automatic segmentation of brain MRI in high-dimensional local and non-local feature space based on sparse representation.

PubMed

Khalilzadeh, Mohammad Mahdi; Fatemizadeh, Emad; Behnam, Hamid

2013-06-01

Automatic extraction of the varying regions of magnetic resonance images is required as a prior step in a diagnostic intelligent system. The sparsest representation and high-dimensional feature are provided based on learned dictionary. The classification is done by employing the technique that computes the reconstruction error locally and non-locally of each pixel. The acquired results from the real and simulated images are superior to the best MRI segmentation method with regard to the stability advantages. In addition, it is segmented exactly through a formula taken from the distance and sparse factors. Also, it is done automatically taking sparse factor in unsupervised clustering methods whose results have been improved. Copyright © 2013 Elsevier Inc. All rights reserved.

Supervised and Unsupervised Learning of Multidimensional Acoustic Categories

ERIC Educational Resources Information Center

Goudbeek, Martijn; Swingley, Daniel; Smits, Roel

2009-01-01

Learning to recognize the contrasts of a language-specific phonemic repertoire can be viewed as forming categories in a multidimensional psychophysical space. Research on the learning of distributionally defined visual categories has shown that categories defined over 1 dimension are easy to learn and that learning multidimensional categories is…

Deep learning-based subdivision approach for large scale macromolecules structure recovery from electron cryo tomograms.

PubMed

Xu, Min; Chai, Xiaoqi; Muthakana, Hariank; Liang, Xiaodan; Yang, Ge; Zeev-Ben-Mordehai, Tzviya; Xing, Eric P

2017-07-15

Cellular Electron CryoTomography (CECT) enables 3D visualization of cellular organization at near-native state and in sub-molecular resolution, making it a powerful tool for analyzing structures of macromolecular complexes and their spatial organizations inside single cells. However, high degree of structural complexity together with practical imaging limitations makes the systematic de novo discovery of structures within cells challenging. It would likely require averaging and classifying millions of subtomograms potentially containing hundreds of highly heterogeneous structural classes. Although it is no longer difficult to acquire CECT data containing such amount of subtomograms due to advances in data acquisition automation, existing computational approaches have very limited scalability or discrimination ability, making them incapable of processing such amount of data. To complement existing approaches, in this article we propose a new approach for subdividing subtomograms into smaller but relatively homogeneous subsets. The structures in these subsets can then be separately recovered using existing computation intensive methods. Our approach is based on supervised structural feature extraction using deep learning, in combination with unsupervised clustering and reference-free classification. Our experiments show that, compared with existing unsupervised rotation invariant feature and pose-normalization based approaches, our new approach achieves significant improvements in both discrimination ability and scalability. More importantly, our new approach is able to discover new structural classes and recover structures that do not exist in training data. Source code freely available at http://www.cs.cmu.edu/∼mxu1/software . mxu1@cs.cmu.edu. Supplementary data are available at Bioinformatics online. © The Author 2017. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com

Machinery running state identification based on discriminant semi-supervised local tangent space alignment for feature fusion and extraction

NASA Astrophysics Data System (ADS)

Su, Zuqiang; Xiao, Hong; Zhang, Yi; Tang, Baoping; Jiang, Yonghua

2017-04-01

Extraction of sensitive features is a challenging but key task in data-driven machinery running state identification. Aimed at solving this problem, a method for machinery running state identification that applies discriminant semi-supervised local tangent space alignment (DSS-LTSA) for feature fusion and extraction is proposed. Firstly, in order to extract more distinct features, the vibration signals are decomposed by wavelet packet decomposition WPD, and a mixed-domain feature set consisted of statistical features, autoregressive (AR) model coefficients, instantaneous amplitude Shannon entropy and WPD energy spectrum is extracted to comprehensively characterize the properties of machinery running state(s). Then, the mixed-dimension feature set is inputted into DSS-LTSA for feature fusion and extraction to eliminate redundant information and interference noise. The proposed DSS-LTSA can extract intrinsic structure information of both labeled and unlabeled state samples, and as a result the over-fitting problem of supervised manifold learning and blindness problem of unsupervised manifold learning are overcome. Simultaneously, class discrimination information is integrated within the dimension reduction process in a semi-supervised manner to improve sensitivity of the extracted fusion features. Lastly, the extracted fusion features are inputted into a pattern recognition algorithm to achieve the running state identification. The effectiveness of the proposed method is verified by a running state identification case in a gearbox, and the results confirm the improved accuracy of the running state identification.

An unsupervised classification approach for analysis of Landsat data to monitor land reclamation in Belmont county, Ohio

NASA Technical Reports Server (NTRS)

Brumfield, J. O.; Bloemer, H. H. L.; Campbell, W. J.

1981-01-01

Two unsupervised classification procedures for analyzing Landsat data used to monitor land reclamation in a surface mining area in east central Ohio are compared for agreement with data collected from the corresponding locations on the ground. One procedure is based on a traditional unsupervised-clustering/maximum-likelihood algorithm sequence that assumes spectral groupings in the Landsat data in n-dimensional space; the other is based on a nontraditional unsupervised-clustering/canonical-transformation/clustering algorithm sequence that not only assumes spectral groupings in n-dimensional space but also includes an additional feature-extraction technique. It is found that the nontraditional procedure provides an appreciable improvement in spectral groupings and apparently increases the level of accuracy in the classification of land cover categories.

Perception Evolution Network Based on Cognition Deepening Model--Adapting to the Emergence of New Sensory Receptor.

PubMed

Xing, Youlu; Shen, Furao; Zhao, Jinxi

2016-03-01

The proposed perception evolution network (PEN) is a biologically inspired neural network model for unsupervised learning and online incremental learning. It is able to automatically learn suitable prototypes from learning data in an incremental way, and it does not require the predefined prototype number or the predefined similarity threshold. Meanwhile, being more advanced than the existing unsupervised neural network model, PEN permits the emergence of a new dimension of perception in the perception field of the network. When a new dimension of perception is introduced, PEN is able to integrate the new dimensional sensory inputs with the learned prototypes, i.e., the prototypes are mapped to a high-dimensional space, which consists of both the original dimension and the new dimension of the sensory inputs. In the experiment, artificial data and real-world data are used to test the proposed PEN, and the results show that PEN can work effectively.

Resting-State fMRI Activity Predicts Unsupervised Learning and Memory in an Immersive Virtual Reality Environment

PubMed Central

Wong, Chi Wah; Olafsson, Valur; Plank, Markus; Snider, Joseph; Halgren, Eric; Poizner, Howard; Liu, Thomas T.

2014-01-01

In the real world, learning often proceeds in an unsupervised manner without explicit instructions or feedback. In this study, we employed an experimental paradigm in which subjects explored an immersive virtual reality environment on each of two days. On day 1, subjects implicitly learned the location of 39 objects in an unsupervised fashion. On day 2, the locations of some of the objects were changed, and object location recall performance was assessed and found to vary across subjects. As prior work had shown that functional magnetic resonance imaging (fMRI) measures of resting-state brain activity can predict various measures of brain performance across individuals, we examined whether resting-state fMRI measures could be used to predict object location recall performance. We found a significant correlation between performance and the variability of the resting-state fMRI signal in the basal ganglia, hippocampus, amygdala, thalamus, insula, and regions in the frontal and temporal lobes, regions important for spatial exploration, learning, memory, and decision making. In addition, performance was significantly correlated with resting-state fMRI connectivity between the left caudate and the right fusiform gyrus, lateral occipital complex, and superior temporal gyrus. Given the basal ganglia's role in exploration, these findings suggest that tighter integration of the brain systems responsible for exploration and visuospatial processing may be critical for learning in a complex environment. PMID:25286145

Unsupervised visual discrimination learning of complex stimuli: Accuracy, bias and generalization.

PubMed

Montefusco-Siegmund, Rodrigo; Toro, Mauricio; Maldonado, Pedro E; Aylwin, María de la L

2018-07-01

Through same-different judgements, we can discriminate an immense variety of stimuli and consequently, they are critical in our everyday interaction with the environment. The quality of the judgements depends on familiarity with stimuli. A way to improve the discrimination is through learning, but to this day, we lack direct evidence of how learning shapes the same-different judgments with complex stimuli. We studied unsupervised visual discrimination learning in 42 participants, as they performed same-different judgments with two types of unfamiliar complex stimuli in the absence of labeling or individuation. Across nine daily training sessions with equiprobable same and different stimuli pairs, participants increased the sensitivity and the criterion by reducing the errors with both same and different pairs. With practice, there was a superior performance for different pairs and a bias for different response. To evaluate the process underlying this bias, we manipulated the proportion of same and different pairs, which resulted in an additional proportion-induced bias, suggesting that the bias observed with equal proportions was a stimulus processing bias. Overall, these results suggest that unsupervised discrimination learning occurs through changes in the stimulus processing that increase the sensory evidence and/or the precision of the working memory. Finally, the acquired discrimination ability was fully transferred to novel exemplars of the practiced stimuli category, in agreement with the acquisition of a category specific perceptual expertise. Copyright © 2018 Elsevier Ltd. All rights reserved.

Incrementally learning objects by touch: online discriminative and generative models for tactile-based recognition.

PubMed

Soh, Harold; Demiris, Yiannis

2014-01-01

Human beings not only possess the remarkable ability to distinguish objects through tactile feedback but are further able to improve upon recognition competence through experience. In this work, we explore tactile-based object recognition with learners capable of incremental learning. Using the sparse online infinite Echo-State Gaussian process (OIESGP), we propose and compare two novel discriminative and generative tactile learners that produce probability distributions over objects during object grasping/palpation. To enable iterative improvement, our online methods incorporate training samples as they become available. We also describe incremental unsupervised learning mechanisms, based on novelty scores and extreme value theory, when teacher labels are not available. We present experimental results for both supervised and unsupervised learning tasks using the iCub humanoid, with tactile sensors on its five-fingered anthropomorphic hand, and 10 different object classes. Our classifiers perform comparably to state-of-the-art methods (C4.5 and SVM classifiers) and findings indicate that tactile signals are highly relevant for making accurate object classifications. We also show that accurate "early" classifications are possible using only 20-30 percent of the grasp sequence. For unsupervised learning, our methods generate high quality clusterings relative to the widely-used sequential k-means and self-organising map (SOM), and we present analyses into the differences between the approaches.

Function approximation using combined unsupervised and supervised learning.

PubMed

Andras, Peter

2014-03-01

Function approximation is one of the core tasks that are solved using neural networks in the context of many engineering problems. However, good approximation results need good sampling of the data space, which usually requires exponentially increasing volume of data as the dimensionality of the data increases. At the same time, often the high-dimensional data is arranged around a much lower dimensional manifold. Here we propose the breaking of the function approximation task for high-dimensional data into two steps: (1) the mapping of the high-dimensional data onto a lower dimensional space corresponding to the manifold on which the data resides and (2) the approximation of the function using the mapped lower dimensional data. We use over-complete self-organizing maps (SOMs) for the mapping through unsupervised learning, and single hidden layer neural networks for the function approximation through supervised learning. We also extend the two-step procedure by considering support vector machines and Bayesian SOMs for the determination of the best parameters for the nonlinear neurons in the hidden layer of the neural networks used for the function approximation. We compare the approximation performance of the proposed neural networks using a set of functions and show that indeed the neural networks using combined unsupervised and supervised learning outperform in most cases the neural networks that learn the function approximation using the original high-dimensional data.

Unsupervised Feature Selection Based on the Morisita Index for Hyperspectral Images

NASA Astrophysics Data System (ADS)

Golay, Jean; Kanevski, Mikhail

2017-04-01

Hyperspectral sensors are capable of acquiring images with hundreds of narrow and contiguous spectral bands. Compared with traditional multispectral imagery, the use of hyperspectral images allows better performance in discriminating between land-cover classes, but it also results in large redundancy and high computational data processing. To alleviate such issues, unsupervised feature selection techniques for redundancy minimization can be implemented. Their goal is to select the smallest subset of features (or bands) in such a way that all the information content of a data set is preserved as much as possible. The present research deals with the application to hyperspectral images of a recently introduced technique of unsupervised feature selection: the Morisita-Based filter for Redundancy Minimization (MBRM). MBRM is based on the (multipoint) Morisita index of clustering and on the Morisita estimator of Intrinsic Dimension (ID). The fundamental idea of the technique is to retain only the bands which contribute to increasing the ID of an image. In this way, redundant bands are disregarded, since they have no impact on the ID. Besides, MBRM has several advantages over benchmark techniques: in addition to its ability to deal with large data sets, it can capture highly-nonlinear dependences and its implementation is straightforward in any programming environment. Experimental results on freely available hyperspectral images show the good effectiveness of MBRM in remote sensing data processing. Comparisons with benchmark techniques are carried out and random forests are used to assess the performance of MBRM in reducing the data dimensionality without loss of relevant information. References [1] C. Traina Jr., A.J.M. Traina, L. Wu, C. Faloutsos, Fast feature selection using fractal dimension, in: Proceedings of the XV Brazilian Symposium on Databases, SBBD, pp. 158-171, 2000. [2] J. Golay, M. Kanevski, A new estimator of intrinsic dimension based on the multipoint Morisita index, Pattern Recognition 48(12), pp. 4070-4081, 2015. [3] J. Golay, M. Kanevski, Unsupervised feature selection based on the Morisita estimator of intrinsic dimension, arXiv:1608.05581, 2016.

Integrating dynamic stopping, transfer learning and language models in an adaptive zero-training ERP speller.

PubMed

Kindermans, Pieter-Jan; Tangermann, Michael; Müller, Klaus-Robert; Schrauwen, Benjamin

2014-06-01

Most BCIs have to undergo a calibration session in which data is recorded to train decoders with machine learning. Only recently zero-training methods have become a subject of study. This work proposes a probabilistic framework for BCI applications which exploit event-related potentials (ERPs). For the example of a visual P300 speller we show how the framework harvests the structure suitable to solve the decoding task by (a) transfer learning, (b) unsupervised adaptation, (c) language model and (d) dynamic stopping. A simulation study compares the proposed probabilistic zero framework (using transfer learning and task structure) to a state-of-the-art supervised model on n = 22 subjects. The individual influence of the involved components (a)-(d) are investigated. Without any need for a calibration session, the probabilistic zero-training framework with inter-subject transfer learning shows excellent performance--competitive to a state-of-the-art supervised method using calibration. Its decoding quality is carried mainly by the effect of transfer learning in combination with continuous unsupervised adaptation. A high-performing zero-training BCI is within reach for one of the most popular BCI paradigms: ERP spelling. Recording calibration data for a supervised BCI would require valuable time which is lost for spelling. The time spent on calibration would allow a novel user to spell 29 symbols with our unsupervised approach. It could be of use for various clinical and non-clinical ERP-applications of BCI.

Integrating dynamic stopping, transfer learning and language models in an adaptive zero-training ERP speller

NASA Astrophysics Data System (ADS)

Kindermans, Pieter-Jan; Tangermann, Michael; Müller, Klaus-Robert; Schrauwen, Benjamin

2014-06-01

Objective. Most BCIs have to undergo a calibration session in which data is recorded to train decoders with machine learning. Only recently zero-training methods have become a subject of study. This work proposes a probabilistic framework for BCI applications which exploit event-related potentials (ERPs). For the example of a visual P300 speller we show how the framework harvests the structure suitable to solve the decoding task by (a) transfer learning, (b) unsupervised adaptation, (c) language model and (d) dynamic stopping. Approach. A simulation study compares the proposed probabilistic zero framework (using transfer learning and task structure) to a state-of-the-art supervised model on n = 22 subjects. The individual influence of the involved components (a)-(d) are investigated. Main results. Without any need for a calibration session, the probabilistic zero-training framework with inter-subject transfer learning shows excellent performance—competitive to a state-of-the-art supervised method using calibration. Its decoding quality is carried mainly by the effect of transfer learning in combination with continuous unsupervised adaptation. Significance. A high-performing zero-training BCI is within reach for one of the most popular BCI paradigms: ERP spelling. Recording calibration data for a supervised BCI would require valuable time which is lost for spelling. The time spent on calibration would allow a novel user to spell 29 symbols with our unsupervised approach. It could be of use for various clinical and non-clinical ERP-applications of BCI.

Analysing exoplanetary data using unsupervised machine-learning

NASA Astrophysics Data System (ADS)

Waldmann, I. P.

2012-04-01

The field of transiting extrasolar planets and especially the study of their atmospheres is one of the youngest and most dynamic subjects in current astrophysics. Permanently at the edge of technical feasibility, we are successfully discovering and characterising smaller and smaller planets. To study exoplanetary atmospheres, we typically require a 10-4 to 10-5 level of accuracy in flux. Achieving such a precision has become the central challenge to exoplanetary research and is often impeded by systematic (nongaussian) noise from either the instrument, stellar activity or both. Dedicated missions, such as Kepler, feature an a priori instrument calibration plan to the required accuracy but nonetheless remain limited by stellar systematics. More generic instruments often lack a sufficiently defined instrument response function, making it very hard to calibrate. In these cases, it becomes interesting to know how well we can calibrate the data without any additional or prior knowledge of the instrument or star. In this conference, we present a non-parametric machine-learning algorithm, based on the concept of independent component analysis, to de-convolve the systematic noise and all non-Gaussian signals from the desired astrophysical signal. Such a 'blind' signal de-mixing is commonly known as the 'Cocktail Party problem' in signal-processing. We showcase the importance and broad applicability of unsupervised machine learning in exoplanetary data analysis by discussing: 1) the removal of instrument systematics in a re-analysis of an HD189733b transmission spectrum obtained with Hubble/NICMOS; 2) the removal of time-correlated stellar noise in individual lightcurves observed by the Kepler mission.

Object-Location-Aware Hashing for Multi-Label Image Retrieval via Automatic Mask Learning.

PubMed

Huang, Chang-Qin; Yang, Shang-Ming; Pan, Yan; Lai, Han-Jiang

2018-09-01

Learning-based hashing is a leading approach of approximate nearest neighbor search for large-scale image retrieval. In this paper, we develop a deep supervised hashing method for multi-label image retrieval, in which we propose to learn a binary "mask" map that can identify the approximate locations of objects in an image, so that we use this binary "mask" map to obtain length-limited hash codes which mainly focus on an image's objects but ignore the background. The proposed deep architecture consists of four parts: 1) a convolutional sub-network to generate effective image features; 2) a binary "mask" sub-network to identify image objects' approximate locations; 3) a weighted average pooling operation based on the binary "mask" to obtain feature representations and hash codes that pay most attention to foreground objects but ignore the background; and 4) the combination of a triplet ranking loss designed to preserve relative similarities among images and a cross entropy loss defined on image labels. We conduct comprehensive evaluations on four multi-label image data sets. The results indicate that the proposed hashing method achieves superior performance gains over the state-of-the-art supervised or unsupervised hashing baselines.

Respiratory Artefact Removal in Forced Oscillation Measurements: A Machine Learning Approach.

PubMed

Pham, Thuy T; Thamrin, Cindy; Robinson, Paul D; McEwan, Alistair L; Leong, Philip H W

2017-08-01

Respiratory artefact removal for the forced oscillation technique can be treated as an anomaly detection problem. Manual removal is currently considered the gold standard, but this approach is laborious and subjective. Most existing automated techniques used simple statistics and/or rejected anomalous data points. Unfortunately, simple statistics are insensitive to numerous artefacts, leading to low reproducibility of results. Furthermore, rejecting anomalous data points causes an imbalance between the inspiratory and expiratory contributions. From a machine learning perspective, such methods are unsupervised and can be considered simple feature extraction. We hypothesize that supervised techniques can be used to find improved features that are more discriminative and more highly correlated with the desired output. Features thus found are then used for anomaly detection by applying quartile thresholding, which rejects complete breaths if one of its features is out of range. The thresholds are determined by both saliency and performance metrics rather than qualitative assumptions as in previous works. Feature ranking indicates that our new landmark features are among the highest scoring candidates regardless of age across saliency criteria. F1-scores, receiver operating characteristic, and variability of the mean resistance metrics show that the proposed scheme outperforms previous simple feature extraction approaches. Our subject-independent detector, 1IQR-SU, demonstrated approval rates of 80.6% for adults and 98% for children, higher than existing methods. Our new features are more relevant. Our removal is objective and comparable to the manual method. This is a critical work to automate forced oscillation technique quality control.

Classification of highly unbalanced CYP450 data of drugs using cost sensitive machine learning techniques.

PubMed

Eitrich, T; Kless, A; Druska, C; Meyer, W; Grotendorst, J

2007-01-01

In this paper, we study the classifications of unbalanced data sets of drugs. As an example we chose a data set of 2D6 inhibitors of cytochrome P450. The human cytochrome P450 2D6 isoform plays a key role in the metabolism of many drugs in the preclinical drug discovery process. We have collected a data set from annotated public data and calculated physicochemical properties with chemoinformatics methods. On top of this data, we have built classifiers based on machine learning methods. Data sets with different class distributions lead to the effect that conventional machine learning methods are biased toward the larger class. To overcome this problem and to obtain sensitive but also accurate classifiers we combine machine learning and feature selection methods with techniques addressing the problem of unbalanced classification, such as oversampling and threshold moving. We have used our own implementation of a support vector machine algorithm as well as the maximum entropy method. Our feature selection is based on the unsupervised McCabe method. The classification results from our test set are compared structurally with compounds from the training set. We show that the applied algorithms enable the effective high throughput in silico classification of potential drug candidates.

Perceptual Learning via Modification of Cortical Top-Down Signals

PubMed Central

Schäfer, Roland; Vasilaki, Eleni; Senn, Walter

2007-01-01

The primary visual cortex (V1) is pre-wired to facilitate the extraction of behaviorally important visual features. Collinear edge detectors in V1, for instance, mutually enhance each other to improve the perception of lines against a noisy background. The same pre-wiring that facilitates line extraction, however, is detrimental when subjects have to discriminate the brightness of different line segments. How is it possible to improve in one task by unsupervised practicing, without getting worse in the other task? The classical view of perceptual learning is that practicing modulates the feedforward input stream through synaptic modifications onto or within V1. However, any rewiring of V1 would deteriorate other perceptual abilities different from the trained one. We propose a general neuronal model showing that perceptual learning can modulate top-down input to V1 in a task-specific way while feedforward and lateral pathways remain intact. Consistent with biological data, the model explains how context-dependent brightness discrimination is improved by a top-down recruitment of recurrent inhibition and a top-down induced increase of the neuronal gain within V1. Both the top-down modulation of inhibition and of neuronal gain are suggested to be universal features of cortical microcircuits which enable perceptual learning. PMID:17715996

Evaluation of solar angle variation over digital processing of LANDSAT imagery. [Brazil

NASA Technical Reports Server (NTRS)

Parada, N. D. J. (Principal Investigator); Novo, E. M. L. M.

1984-01-01

The effects of the seasonal variation of illumination over digital processing of LANDSAT images are evaluated. Original images are transformed by means of digital filtering to enhance their spatial features. The resulting images are used to obtain an unsupervised classification of relief units. After defining relief classes, which are supposed to be spectrally different, topographic variables (declivity, altitude, relief range and slope length) are used to identify the true relief units existing on the ground. The samples are also clustered by means of an unsupervised classification option. The results obtained for each LANDSAT overpass are compared. Digital processing is highly affected by illumination geometry. There is no correspondence between relief units as defined by spectral features and those resulting from topographic features.

Discriminative clustering on manifold for adaptive transductive classification.

PubMed

Zhang, Zhao; Jia, Lei; Zhang, Min; Li, Bing; Zhang, Li; Li, Fanzhang

2017-10-01

In this paper, we mainly propose a novel adaptive transductive label propagation approach by joint discriminative clustering on manifolds for representing and classifying high-dimensional data. Our framework seamlessly combines the unsupervised manifold learning, discriminative clustering and adaptive classification into a unified model. Also, our method incorporates the adaptive graph weight construction with label propagation. Specifically, our method is capable of propagating label information using adaptive weights over low-dimensional manifold features, which is different from most existing studies that usually predict the labels and construct the weights in the original Euclidean space. For transductive classification by our formulation, we first perform the joint discriminative K-means clustering and manifold learning to capture the low-dimensional nonlinear manifolds. Then, we construct the adaptive weights over the learnt manifold features, where the adaptive weights are calculated through performing the joint minimization of the reconstruction errors over features and soft labels so that the graph weights can be joint-optimal for data representation and classification. Using the adaptive weights, we can easily estimate the unknown labels of samples. After that, our method returns the updated weights for further updating the manifold features. Extensive simulations on image classification and segmentation show that our proposed algorithm can deliver the state-of-the-art performance on several public datasets. Copyright © 2017 Elsevier Ltd. All rights reserved.

Natural-Annotation-based Unsupervised Construction of Korean-Chinese Domain Dictionary

NASA Astrophysics Data System (ADS)

Liu, Wuying; Wang, Lin

2018-03-01

The large-scale bilingual parallel resource is significant to statistical learning and deep learning in natural language processing. This paper addresses the automatic construction issue of the Korean-Chinese domain dictionary, and presents a novel unsupervised construction method based on the natural annotation in the raw corpus. We firstly extract all Korean-Chinese word pairs from Korean texts according to natural annotations, secondly transform the traditional Chinese characters into the simplified ones, and finally distill out a bilingual domain dictionary after retrieving the simplified Chinese words in an extra Chinese domain dictionary. The experimental results show that our method can automatically build multiple Korean-Chinese domain dictionaries efficiently.

Unsupervised Learning Through Randomized Algorithms for High-Volume High-Velocity Data (ULTRA-HV).

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

Pinar, Ali; Kolda, Tamara G.; Carlberg, Kevin Thomas

Through long-term investments in computing, algorithms, facilities, and instrumentation, DOE is an established leader in massive-scale, high-fidelity simulations, as well as science-leading experimentation. In both cases, DOE is generating more data than it can analyze and the problem is intensifying quickly. The need for advanced algorithms that can automatically convert the abundance of data into a wealth of useful information by discovering hidden structures is well recognized. Such efforts however, are hindered by the massive volume of the data and its high velocity. Here, the challenge is developing unsupervised learning methods to discover hidden structure in high-volume, high-velocity data.

Full-body gestures and movements recognition: user descriptive and unsupervised learning approaches in GDL classifier

NASA Astrophysics Data System (ADS)

Hachaj, Tomasz; Ogiela, Marek R.

2014-09-01

Gesture Description Language (GDL) is a classifier that enables syntactic description and real time recognition of full-body gestures and movements. Gestures are described in dedicated computer language named Gesture Description Language script (GDLs). In this paper we will introduce new GDLs formalisms that enable recognition of selected classes of movement trajectories. The second novelty is new unsupervised learning method with which it is possible to automatically generate GDLs descriptions. We have initially evaluated both proposed extensions of GDL and we have obtained very promising results. Both the novel methodology and evaluation results will be described in this paper.

Learning about individuals' health from aggregate data.

PubMed

Colbaugh, Rich; Glass, Kristin

2017-07-01

There is growing awareness that user-generated social media content contains valuable health-related information and is more convenient to collect than typical health data. For example, Twitter has been employed to predict aggregate-level outcomes, such as regional rates of diabetes and child poverty, and to identify individual cases of depression and food poisoning. Models which make aggregate-level inferences can be induced from aggregate data, and consequently are straightforward to build. In contrast, learning models that produce individual-level (IL) predictions, which are more informative, usually requires a large number of difficult-to-acquire labeled IL examples. This paper presents a new machine learning method which achieves the best of both worlds, enabling IL models to be learned from aggregate labels. The algorithm makes predictions by combining unsupervised feature extraction, aggregate-based modeling, and optimal integration of aggregate-level and IL information. Two case studies illustrate how to learn health-relevant IL prediction models using only aggregate labels, and show that these models perform as well as state-of-the-art models trained on hundreds or thousands of labeled individuals.

A comparison of neural network and fuzzy clustering techniques in segmenting magnetic resonance images of the brain.

PubMed

Hall, L O; Bensaid, A M; Clarke, L P; Velthuizen, R P; Silbiger, M S; Bezdek, J C

1992-01-01

Magnetic resonance (MR) brain section images are segmented and then synthetically colored to give visual representations of the original data with three approaches: the literal and approximate fuzzy c-means unsupervised clustering algorithms, and a supervised computational neural network. Initial clinical results are presented on normal volunteers and selected patients with brain tumors surrounded by edema. Supervised and unsupervised segmentation techniques provide broadly similar results. Unsupervised fuzzy algorithms were visually observed to show better segmentation when compared with raw image data for volunteer studies. For a more complex segmentation problem with tumor/edema or cerebrospinal fluid boundary, where the tissues have similar MR relaxation behavior, inconsistency in rating among experts was observed, with fuzz-c-means approaches being slightly preferred over feedforward cascade correlation results. Various facets of both approaches, such as supervised versus unsupervised learning, time complexity, and utility for the diagnostic process, are compared.

Identification and tracking of vertebrae in ultrasound using deep networks with unsupervised feature learning

NASA Astrophysics Data System (ADS)

Hetherington, Jorden; Pesteie, Mehran; Lessoway, Victoria A.; Abolmaesumi, Purang; Rohling, Robert N.

2017-03-01

Percutaneous needle insertion procedures on the spine often require proper identification of the vertebral level in order to effectively deliver anesthetics and analgesic agents to achieve adequate block. For example, in obstetric epidurals, the target is at the L3-L4 intervertebral space. The current clinical method involves "blind" identification of the vertebral level through manual palpation of the spine, which has only 30% accuracy. This implies the need for better anatomical identification prior to needle insertion. A system is proposed to identify the vertebrae, assigning them to their respective levels, and track them in a standard sequence of ultrasound images, when imaged in the paramedian plane. Machine learning techniques are developed to identify discriminative features of the laminae. In particular, a deep network is trained to automatically learn the anatomical features of the lamina peaks, and classify image patches, for pixel-level classification. The chosen network utilizes multiple connected auto-encoders to learn the anatomy. Pre-processing with ultrasound bone enhancement techniques is done to aid the pixel-level classification performance. Once the lamina are identified, vertebrae are assigned levels and tracked in sequential frames. Experimental results were evaluated against an expert sonographer. Based on data acquired from 15 subjects, vertebrae identification with sensitivity of 95% and precision of 95% was achieved within each frame. Between pairs of subsequently analyzed frames, matches of predicted vertebral level labels were correct in 94% of cases, when compared to matches of manually selected labels

Adding Learning to Knowledge-Based Systems: Taking the "Artificial" Out of AI

Treesearch

Daniel L. Schmoldt

1997-01-01

Both, knowledge-based systems (KBS) development and maintenance require time-consuming analysis of domain knowledge. Where example cases exist, KBS can be built, and later updated, by incorporating learning capabilities into their architecture. This applies to both supervised and unsupervised learning scenarios. In this paper, the important issues for learning systems-...

Author Detection on a Mobile Phone

DTIC Science & Technology

2011-03-01

handwriting , and to mine sales data for profitable trends. Two broad categories of machine learning are supervised learn- ing and unsupervised learning...evaluation,” AI 2006: Advances in Artificial Intelligence, p. 1015–1021, 2006. [23] “Gartner says worldwide mobile phone sales grew 17 per cent in first

Strong systematicity through sensorimotor conceptual grounding: an unsupervised, developmental approach to connectionist sentence processing

NASA Astrophysics Data System (ADS)

Jansen, Peter A.; Watter, Scott

2012-03-01

Connectionist language modelling typically has difficulty with syntactic systematicity, or the ability to generalise language learning to untrained sentences. This work develops an unsupervised connectionist model of infant grammar learning. Following the semantic boostrapping hypothesis, the network distils word category using a developmentally plausible infant-scale database of grounded sensorimotor conceptual representations, as well as a biologically plausible semantic co-occurrence activation function. The network then uses this knowledge to acquire an early benchmark clausal grammar using correlational learning, and further acquires separate conceptual and grammatical category representations. The network displays strongly systematic behaviour indicative of the general acquisition of the combinatorial systematicity present in the grounded infant-scale language stream, outperforms previous contemporary models that contain primarily noun and verb word categories, and successfully generalises broadly to novel untrained sensorimotor grounded sentences composed of unfamiliar nouns and verbs. Limitations as well as implications to later grammar learning are discussed.

An unsupervised technique for optimal feature selection in attribute profiles for spectral-spatial classification of hyperspectral images

NASA Astrophysics Data System (ADS)

Bhardwaj, Kaushal; Patra, Swarnajyoti

2018-04-01

Inclusion of spatial information along with spectral features play a significant role in classification of remote sensing images. Attribute profiles have already proved their ability to represent spatial information. In order to incorporate proper spatial information, multiple attributes are required and for each attribute large profiles need to be constructed by varying the filter parameter values within a wide range. Thus, the constructed profiles that represent spectral-spatial information of an hyperspectral image have huge dimension which leads to Hughes phenomenon and increases computational burden. To mitigate these problems, this work presents an unsupervised feature selection technique that selects a subset of filtered image from the constructed high dimensional multi-attribute profile which are sufficiently informative to discriminate well among classes. In this regard the proposed technique exploits genetic algorithms (GAs). The fitness function of GAs are defined in an unsupervised way with the help of mutual information. The effectiveness of the proposed technique is assessed using one-against-all support vector machine classifier. The experiments conducted on three hyperspectral data sets show the robustness of the proposed method in terms of computation time and classification accuracy.

The oligonucleotide frequency derived error gradient and its application to the binning of metagenome fragments

PubMed Central

2009-01-01

Background The characterisation, or binning, of metagenome fragments is an important first step to further downstream analysis of microbial consortia. Here, we propose a one-dimensional signature, OFDEG, derived from the oligonucleotide frequency profile of a DNA sequence, and show that it is possible to obtain a meaningful phylogenetic signal for relatively short DNA sequences. The one-dimensional signal is essentially a compact representation of higher dimensional feature spaces of greater complexity and is intended to improve on the tetranucleotide frequency feature space preferred by current compositional binning methods. Results We compare the fidelity of OFDEG against tetranucleotide frequency in both an unsupervised and semi-supervised setting on simulated metagenome benchmark data. Four tests were conducted using assembler output of Arachne and phrap, and for each, performance was evaluated on contigs which are greater than or equal to 8 kbp in length and contigs which are composed of at least 10 reads. Using both G-C content in conjunction with OFDEG gave an average accuracy of 96.75% (semi-supervised) and 95.19% (unsupervised), versus 94.25% (semi-supervised) and 82.35% (unsupervised) for tetranucleotide frequency. Conclusion We have presented an observation of an alternative characteristic of DNA sequences. The proposed feature representation has proven to be more beneficial than the existing tetranucleotide frequency space to the metagenome binning problem. We do note, however, that our observation of OFDEG deserves further anlaysis and investigation. Unsupervised clustering revealed OFDEG related features performed better than standard tetranucleotide frequency in representing a relevant organism specific signal. Further improvement in binning accuracy is given by semi-supervised classification using OFDEG. The emphasis on a feature-driven, bottom-up approach to the problem of binning reveals promising avenues for future development of techniques to characterise short environmental sequences without bias toward cultivable organisms. PMID:19958473

Unsupervised classification of major depression using functional connectivity MRI.

PubMed

Zeng, Ling-Li; Shen, Hui; Liu, Li; Hu, Dewen

2014-04-01

The current diagnosis of psychiatric disorders including major depressive disorder based largely on self-reported symptoms and clinical signs may be prone to patients' behaviors and psychiatrists' bias. This study aims at developing an unsupervised machine learning approach for the accurate identification of major depression based on single resting-state functional magnetic resonance imaging scans in the absence of clinical information. Twenty-four medication-naive patients with major depression and 29 demographically similar healthy individuals underwent resting-state functional magnetic resonance imaging. We first clustered the voxels within the perigenual cingulate cortex into two subregions, a subgenual region and a pregenual region, according to their distinct resting-state functional connectivity patterns and showed that a maximum margin clustering-based unsupervised machine learning approach extracted sufficient information from the subgenual cingulate functional connectivity map to differentiate depressed patients from healthy controls with a group-level clustering consistency of 92.5% and an individual-level classification consistency of 92.5%. It was also revealed that the subgenual cingulate functional connectivity network with the highest discriminative power primarily included the ventrolateral and ventromedial prefrontal cortex, superior temporal gyri and limbic areas, indicating that these connections may play critical roles in the pathophysiology of major depression. The current study suggests that subgenual cingulate functional connectivity network signatures may provide promising objective biomarkers for the diagnosis of major depression and that maximum margin clustering-based unsupervised machine learning approaches may have the potential to inform clinical practice and aid in research on psychiatric disorders. Copyright © 2013 Wiley Periodicals, Inc.

Discovering phases, phase transitions, and crossovers through unsupervised machine learning: A critical examination

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

Hu, Wenjian; Singh, Rajiv R. P.; Scalettar, Richard T.

Here, we apply unsupervised machine learning techniques, mainly principal component analysis (PCA), to compare and contrast the phase behavior and phase transitions in several classical spin models - the square and triangular-lattice Ising models, the Blume-Capel model, a highly degenerate biquadratic-exchange spin-one Ising (BSI) model, and the 2D XY model, and examine critically what machine learning is teaching us. We find that quantified principal components from PCA not only allow exploration of different phases and symmetry-breaking, but can distinguish phase transition types and locate critical points. We show that the corresponding weight vectors have a clear physical interpretation, which ismore » particularly interesting in the frustrated models such as the triangular antiferromagnet, where they can point to incipient orders. Unlike the other well-studied models, the properties of the BSI model are less well known. Using both PCA and conventional Monte Carlo analysis, we demonstrate that the BSI model shows an absence of phase transition and macroscopic ground-state degeneracy. The failure to capture the 'charge' correlations (vorticity) in the BSI model (XY model) from raw spin configurations points to some of the limitations of PCA. Finally, we employ a nonlinear unsupervised machine learning procedure, the 'antoencoder method', and demonstrate that it too can be trained to capture phase transitions and critical points.« less

Discovering phases, phase transitions, and crossovers through unsupervised machine learning: A critical examination

DOE PAGES

Hu, Wenjian; Singh, Rajiv R. P.; Scalettar, Richard T.

2017-06-19

Here, we apply unsupervised machine learning techniques, mainly principal component analysis (PCA), to compare and contrast the phase behavior and phase transitions in several classical spin models - the square and triangular-lattice Ising models, the Blume-Capel model, a highly degenerate biquadratic-exchange spin-one Ising (BSI) model, and the 2D XY model, and examine critically what machine learning is teaching us. We find that quantified principal components from PCA not only allow exploration of different phases and symmetry-breaking, but can distinguish phase transition types and locate critical points. We show that the corresponding weight vectors have a clear physical interpretation, which ismore » particularly interesting in the frustrated models such as the triangular antiferromagnet, where they can point to incipient orders. Unlike the other well-studied models, the properties of the BSI model are less well known. Using both PCA and conventional Monte Carlo analysis, we demonstrate that the BSI model shows an absence of phase transition and macroscopic ground-state degeneracy. The failure to capture the 'charge' correlations (vorticity) in the BSI model (XY model) from raw spin configurations points to some of the limitations of PCA. Finally, we employ a nonlinear unsupervised machine learning procedure, the 'antoencoder method', and demonstrate that it too can be trained to capture phase transitions and critical points.« less

Discovering phases, phase transitions, and crossovers through unsupervised machine learning: A critical examination

NASA Astrophysics Data System (ADS)

Hu, Wenjian; Singh, Rajiv R. P.; Scalettar, Richard T.

2017-06-01

We apply unsupervised machine learning techniques, mainly principal component analysis (PCA), to compare and contrast the phase behavior and phase transitions in several classical spin models—the square- and triangular-lattice Ising models, the Blume-Capel model, a highly degenerate biquadratic-exchange spin-1 Ising (BSI) model, and the two-dimensional X Y model—and we examine critically what machine learning is teaching us. We find that quantified principal components from PCA not only allow the exploration of different phases and symmetry-breaking, but they can distinguish phase-transition types and locate critical points. We show that the corresponding weight vectors have a clear physical interpretation, which is particularly interesting in the frustrated models such as the triangular antiferromagnet, where they can point to incipient orders. Unlike the other well-studied models, the properties of the BSI model are less well known. Using both PCA and conventional Monte Carlo analysis, we demonstrate that the BSI model shows an absence of phase transition and macroscopic ground-state degeneracy. The failure to capture the "charge" correlations (vorticity) in the BSI model (X Y model) from raw spin configurations points to some of the limitations of PCA. Finally, we employ a nonlinear unsupervised machine learning procedure, the "autoencoder method," and we demonstrate that it too can be trained to capture phase transitions and critical points.

Unsupervised segmentation of lung fields in chest radiographs using multiresolution fractal feature vector and deformable models.

PubMed

Lee, Wen-Li; Chang, Koyin; Hsieh, Kai-Sheng

2016-09-01

Segmenting lung fields in a chest radiograph is essential for automatically analyzing an image. We present an unsupervised method based on multiresolution fractal feature vector. The feature vector characterizes the lung field region effectively. A fuzzy c-means clustering algorithm is then applied to obtain a satisfactory initial contour. The final contour is obtained by deformable models. The results show the feasibility and high performance of the proposed method. Furthermore, based on the segmentation of lung fields, the cardiothoracic ratio (CTR) can be measured. The CTR is a simple index for evaluating cardiac hypertrophy. After identifying a suspicious symptom based on the estimated CTR, a physician can suggest that the patient undergoes additional extensive tests before a treatment plan is finalized.

Modeling language and cognition with deep unsupervised learning: a tutorial overview

PubMed Central

Zorzi, Marco; Testolin, Alberto; Stoianov, Ivilin P.

2013-01-01

Deep unsupervised learning in stochastic recurrent neural networks with many layers of hidden units is a recent breakthrough in neural computation research. These networks build a hierarchy of progressively more complex distributed representations of the sensory data by fitting a hierarchical generative model. In this article we discuss the theoretical foundations of this approach and we review key issues related to training, testing and analysis of deep networks for modeling language and cognitive processing. The classic letter and word perception problem of McClelland and Rumelhart (1981) is used as a tutorial example to illustrate how structured and abstract representations may emerge from deep generative learning. We argue that the focus on deep architectures and generative (rather than discriminative) learning represents a crucial step forward for the connectionist modeling enterprise, because it offers a more plausible model of cortical learning as well as a way to bridge the gap between emergentist connectionist models and structured Bayesian models of cognition. PMID:23970869

Modeling language and cognition with deep unsupervised learning: a tutorial overview.

PubMed

Zorzi, Marco; Testolin, Alberto; Stoianov, Ivilin P

2013-01-01

Deep unsupervised learning in stochastic recurrent neural networks with many layers of hidden units is a recent breakthrough in neural computation research. These networks build a hierarchy of progressively more complex distributed representations of the sensory data by fitting a hierarchical generative model. In this article we discuss the theoretical foundations of this approach and we review key issues related to training, testing and analysis of deep networks for modeling language and cognitive processing. The classic letter and word perception problem of McClelland and Rumelhart (1981) is used as a tutorial example to illustrate how structured and abstract representations may emerge from deep generative learning. We argue that the focus on deep architectures and generative (rather than discriminative) learning represents a crucial step forward for the connectionist modeling enterprise, because it offers a more plausible model of cortical learning as well as a way to bridge the gap between emergentist connectionist models and structured Bayesian models of cognition.

Functional requirements for reward-modulated spike-timing-dependent plasticity.

PubMed

Frémaux, Nicolas; Sprekeler, Henning; Gerstner, Wulfram

2010-10-06

Recent experiments have shown that spike-timing-dependent plasticity is influenced by neuromodulation. We derive theoretical conditions for successful learning of reward-related behavior for a large class of learning rules where Hebbian synaptic plasticity is conditioned on a global modulatory factor signaling reward. We show that all learning rules in this class can be separated into a term that captures the covariance of neuronal firing and reward and a second term that presents the influence of unsupervised learning. The unsupervised term, which is, in general, detrimental for reward-based learning, can be suppressed if the neuromodulatory signal encodes the difference between the reward and the expected reward-but only if the expected reward is calculated for each task and stimulus separately. If several tasks are to be learned simultaneously, the nervous system needs an internal critic that is able to predict the expected reward for arbitrary stimuli. We show that, with a critic, reward-modulated spike-timing-dependent plasticity is capable of learning motor trajectories with a temporal resolution of tens of milliseconds. The relation to temporal difference learning, the relevance of block-based learning paradigms, and the limitations of learning with a critic are discussed.

Creating Turbulent Flow Realizations with Generative Adversarial Networks

NASA Astrophysics Data System (ADS)

King, Ryan; Graf, Peter; Chertkov, Michael

2017-11-01

Generating valid inflow conditions is a crucial, yet computationally expensive, step in unsteady turbulent flow simulations. We demonstrate a new technique for rapid generation of turbulent inflow realizations that leverages recent advances in machine learning for image generation using a deep convolutional generative adversarial network (DCGAN). The DCGAN is an unsupervised machine learning technique consisting of two competing neural networks that are trained against each other using backpropagation. One network, the generator, tries to produce samples from the true distribution of states, while the discriminator tries to distinguish between true and synthetic samples. We present results from a fully-trained DCGAN that is able to rapidly draw random samples from the full distribution of possible inflow states without needing to solve the Navier-Stokes equations, eliminating the costly process of spinning up inflow turbulence. This suggests a new paradigm in physics informed machine learning where the turbulence physics can be encoded in either the discriminator or generator. Finally, we also propose additional applications such as feature identification and subgrid scale modeling.

Imaging nanoscale lattice variations by machine learning of x-ray diffraction microscopy data

DOE PAGES

Laanait, Nouamane; Zhang, Zhan; Schlepütz, Christian M.

2016-08-09

In this paper, we present a novel methodology based on machine learning to extract lattice variations in crystalline materials, at the nanoscale, from an x-ray Bragg diffraction-based imaging technique. By employing a full-field microscopy setup, we capture real space images of materials, with imaging contrast determined solely by the x-ray diffracted signal. The data sets that emanate from this imaging technique are a hybrid of real space information (image spatial support) and reciprocal lattice space information (image contrast), and are intrinsically multidimensional (5D). By a judicious application of established unsupervised machine learning techniques and multivariate analysis to this multidimensional datamore » cube, we show how to extract features that can be ascribed physical interpretations in terms of common structural distortions, such as lattice tilts and dislocation arrays. Finally, we demonstrate this 'big data' approach to x-ray diffraction microscopy by identifying structural defects present in an epitaxial ferroelectric thin-film of lead zirconate titanate.« less

Imaging nanoscale lattice variations by machine learning of x-ray diffraction microscopy data

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

Laanait, Nouamane; Zhang, Zhan; Schlepütz, Christian M.

In this paper, we present a novel methodology based on machine learning to extract lattice variations in crystalline materials, at the nanoscale, from an x-ray Bragg diffraction-based imaging technique. By employing a full-field microscopy setup, we capture real space images of materials, with imaging contrast determined solely by the x-ray diffracted signal. The data sets that emanate from this imaging technique are a hybrid of real space information (image spatial support) and reciprocal lattice space information (image contrast), and are intrinsically multidimensional (5D). By a judicious application of established unsupervised machine learning techniques and multivariate analysis to this multidimensional datamore » cube, we show how to extract features that can be ascribed physical interpretations in terms of common structural distortions, such as lattice tilts and dislocation arrays. Finally, we demonstrate this 'big data' approach to x-ray diffraction microscopy by identifying structural defects present in an epitaxial ferroelectric thin-film of lead zirconate titanate.« less

Automatic segmentation of amyloid plaques in MR images using unsupervised SVM

PubMed Central

Iordanescu, Gheorghe; Venkatasubramanian, Palamadai N.; Wyrwicz, Alice M.

2011-01-01

Deposition of the β-amyloid peptide (Aβ) is an important pathological hallmark of Alzheimer’s disease (AD). However, reliable quantification of amyloid plaques in both human and animal brains remains a challenge. We present here a novel automatic plaque segmentation algorithm based on the intrinsic MR signal characteristics of plaques. This algorithm identifies plaque candidates in MR data by using watershed transform, which extracts regions with low intensities completely surrounded by higher intensity neighbors. These candidates are classified as plaque or non-plaque by an unsupervised learning method using features derived from the MR data intensity. The algorithm performance is validated by comparison with histology. We also demonstrate the algorithm’s ability to detect age-related changes in plaque load ex vivo in 5×FAD APP transgenic mice. To our knowledge, this work represents the first quantitative method for characterizing amyloid plaques in MRI data. The proposed method can be used to describe the spatio-temporal progression of amyloid deposition, which is necessary for understanding the evolution of plaque pathology in mouse models of AD and to evaluate the efficacy of emergent amyloid-targeting therapies in preclinical trials. PMID:22189675

Mastication Evaluation With Unsupervised Learning: Using an Inertial Sensor-Based System.

PubMed

Lucena, Caroline Vieira; Lacerda, Marcelo; Caldas, Rafael; De Lima Neto, Fernando Buarque; Rativa, Diego

2018-01-01

There is a direct relationship between the prevalence of musculoskeletal disorders of the temporomandibular joint and orofacial disorders. A well-elaborated analysis of the jaw movements provides relevant information for healthcare professionals to conclude their diagnosis. Different approaches have been explored to track jaw movements such that the mastication analysis is getting less subjective; however, all methods are still highly subjective, and the quality of the assessments depends much on the experience of the health professional. In this paper, an accurate and non-invasive method based on a commercial low-cost inertial sensor (MPU6050) to measure jaw movements is proposed. The jaw-movement feature values are compared to the obtained with clinical analysis, showing no statistically significant difference between both methods. Moreover, We propose to use unsupervised paradigm approaches to cluster mastication patterns of healthy subjects and simulated patients with facial trauma. Two techniques were used in this paper to instantiate the method: Kohonen's Self-Organizing Maps and K-Means Clustering. Both algorithms have excellent performances to process jaw-movements data, showing encouraging results and potential to bring a full assessment of the masticatory function. The proposed method can be applied in real-time providing relevant dynamic information for health-care professionals.

Unsupervised EEG analysis for automated epileptic seizure detection

NASA Astrophysics Data System (ADS)

Birjandtalab, Javad; Pouyan, Maziyar Baran; Nourani, Mehrdad

2016-07-01

Epilepsy is a neurological disorder which can, if not controlled, potentially cause unexpected death. It is extremely crucial to have accurate automatic pattern recognition and data mining techniques to detect the onset of seizures and inform care-givers to help the patients. EEG signals are the preferred biosignals for diagnosis of epileptic patients. Most of the existing pattern recognition techniques used in EEG analysis leverage the notion of supervised machine learning algorithms. Since seizure data are heavily under-represented, such techniques are not always practical particularly when the labeled data is not sufficiently available or when disease progression is rapid and the corresponding EEG footprint pattern will not be robust. Furthermore, EEG pattern change is highly individual dependent and requires experienced specialists to annotate the seizure and non-seizure events. In this work, we present an unsupervised technique to discriminate seizures and non-seizures events. We employ power spectral density of EEG signals in different frequency bands that are informative features to accurately cluster seizure and non-seizure events. The experimental results tried so far indicate achieving more than 90% accuracy in clustering seizure and non-seizure events without having any prior knowledge on patient's history.

Scale-invariant feature extraction of neural network and renormalization group flow

NASA Astrophysics Data System (ADS)

Iso, Satoshi; Shiba, Shotaro; Yokoo, Sumito

2018-05-01

Theoretical understanding of how a deep neural network (DNN) extracts features from input images is still unclear, but it is widely believed that the extraction is performed hierarchically through a process of coarse graining. It reminds us of the basic renormalization group (RG) concept in statistical physics. In order to explore possible relations between DNN and RG, we use the restricted Boltzmann machine (RBM) applied to an Ising model and construct a flow of model parameters (in particular, temperature) generated by the RBM. We show that the unsupervised RBM trained by spin configurations at various temperatures from T =0 to T =6 generates a flow along which the temperature approaches the critical value Tc=2.2 7 . This behavior is the opposite of the typical RG flow of the Ising model. By analyzing various properties of the weight matrices of the trained RBM, we discuss why it flows towards Tc and how the RBM learns to extract features of spin configurations.

Subgraph augmented non-negative tensor factorization (SANTF) for modeling clinical narrative text

PubMed Central

Xin, Yu; Hochberg, Ephraim; Joshi, Rohit; Uzuner, Ozlem; Szolovits, Peter

2015-01-01

Objective Extracting medical knowledge from electronic medical records requires automated approaches to combat scalability limitations and selection biases. However, existing machine learning approaches are often regarded by clinicians as black boxes. Moreover, training data for these automated approaches at often sparsely annotated at best. The authors target unsupervised learning for modeling clinical narrative text, aiming at improving both accuracy and interpretability. Methods The authors introduce a novel framework named subgraph augmented non-negative tensor factorization (SANTF). In addition to relying on atomic features (e.g., words in clinical narrative text), SANTF automatically mines higher-order features (e.g., relations of lymphoid cells expressing antigens) from clinical narrative text by converting sentences into a graph representation and identifying important subgraphs. The authors compose a tensor using patients, higher-order features, and atomic features as its respective modes. We then apply non-negative tensor factorization to cluster patients, and simultaneously identify latent groups of higher-order features that link to patient clusters, as in clinical guidelines where a panel of immunophenotypic features and laboratory results are used to specify diagnostic criteria. Results and Conclusion SANTF demonstrated over 10% improvement in averaged F-measure on patient clustering compared to widely used non-negative matrix factorization (NMF) and k-means clustering methods. Multiple baselines were established by modeling patient data using patient-by-features matrices with different feature configurations and then performing NMF or k-means to cluster patients. Feature analysis identified latent groups of higher-order features that lead to medical insights. We also found that the latent groups of atomic features help to better correlate the latent groups of higher-order features. PMID:25862765

Linear time relational prototype based learning.

PubMed

Gisbrecht, Andrej; Mokbel, Bassam; Schleif, Frank-Michael; Zhu, Xibin; Hammer, Barbara

2012-10-01

Prototype based learning offers an intuitive interface to inspect large quantities of electronic data in supervised or unsupervised settings. Recently, many techniques have been extended to data described by general dissimilarities rather than Euclidean vectors, so-called relational data settings. Unlike the Euclidean counterparts, the techniques have quadratic time complexity due to the underlying quadratic dissimilarity matrix. Thus, they are infeasible already for medium sized data sets. The contribution of this article is twofold: On the one hand we propose a novel supervised prototype based classification technique for dissimilarity data based on popular learning vector quantization (LVQ), on the other hand we transfer a linear time approximation technique, the Nyström approximation, to this algorithm and an unsupervised counterpart, the relational generative topographic mapping (GTM). This way, linear time and space methods result. We evaluate the techniques on three examples from the biomedical domain.

ClusterTAD: an unsupervised machine learning approach to detecting topologically associated domains of chromosomes from Hi-C data.

PubMed

Oluwadare, Oluwatosin; Cheng, Jianlin

2017-11-14

With the development of chromosomal conformation capturing techniques, particularly, the Hi-C technique, the study of the spatial conformation of a genome is becoming an important topic in bioinformatics and computational biology. The Hi-C technique can generate genome-wide chromosomal interaction (contact) data, which can be used to investigate the higher-level organization of chromosomes, such as Topologically Associated Domains (TAD), i.e., locally packed chromosome regions bounded together by intra chromosomal contacts. The identification of the TADs for a genome is useful for studying gene regulation, genomic interaction, and genome function. Here, we formulate the TAD identification problem as an unsupervised machine learning (clustering) problem, and develop a new TAD identification method called ClusterTAD. We introduce a novel method to represent chromosomal contacts as features to be used by the clustering algorithm. Our results show that ClusterTAD can accurately predict the TADs on a simulated Hi-C data. Our method is also largely complementary and consistent with existing methods on the real Hi-C datasets of two mouse cells. The validation with the chromatin immunoprecipitation (ChIP) sequencing (ChIP-Seq) data shows that the domain boundaries identified by ClusterTAD have a high enrichment of CTCF binding sites, promoter-related marks, and enhancer-related histone modifications. As ClusterTAD is based on a proven clustering approach, it opens a new avenue to apply a large array of clustering methods developed in the machine learning field to the TAD identification problem. The source code, the results, and the TADs generated for the simulated and real Hi-C datasets are available here: https://github.com/BDM-Lab/ClusterTAD .

Unsupervised discrimination of patterns in spiking neural networks with excitatory and inhibitory synaptic plasticity

PubMed Central

Srinivasa, Narayan; Cho, Youngkwan

2014-01-01

A spiking neural network model is described for learning to discriminate among spatial patterns in an unsupervised manner. The network anatomy consists of source neurons that are activated by external inputs, a reservoir that resembles a generic cortical layer with an excitatory-inhibitory (EI) network and a sink layer of neurons for readout. Synaptic plasticity in the form of STDP is imposed on all the excitatory and inhibitory synapses at all times. While long-term excitatory STDP enables sparse and efficient learning of the salient features in inputs, inhibitory STDP enables this learning to be stable by establishing a balance between excitatory and inhibitory currents at each neuron in the network. The synaptic weights between source and reservoir neurons form a basis set for the input patterns. The neural trajectories generated in the reservoir due to input stimulation and lateral connections between reservoir neurons can be readout by the sink layer neurons. This activity is used for adaptation of synapses between reservoir and sink layer neurons. A new measure called the discriminability index (DI) is introduced to compute if the network can discriminate between old patterns already presented in an initial training session. The DI is also used to compute if the network adapts to new patterns without losing its ability to discriminate among old patterns. The final outcome is that the network is able to correctly discriminate between all patterns—both old and new. This result holds as long as inhibitory synapses employ STDP to continuously enable current balance in the network. The results suggest a possible direction for future investigation into how spiking neural networks could address the stability-plasticity question despite having continuous synaptic plasticity. PMID:25566045

Unsupervised discrimination of patterns in spiking neural networks with excitatory and inhibitory synaptic plasticity.

PubMed

Srinivasa, Narayan; Cho, Youngkwan

2014-01-01

A spiking neural network model is described for learning to discriminate among spatial patterns in an unsupervised manner. The network anatomy consists of source neurons that are activated by external inputs, a reservoir that resembles a generic cortical layer with an excitatory-inhibitory (EI) network and a sink layer of neurons for readout. Synaptic plasticity in the form of STDP is imposed on all the excitatory and inhibitory synapses at all times. While long-term excitatory STDP enables sparse and efficient learning of the salient features in inputs, inhibitory STDP enables this learning to be stable by establishing a balance between excitatory and inhibitory currents at each neuron in the network. The synaptic weights between source and reservoir neurons form a basis set for the input patterns. The neural trajectories generated in the reservoir due to input stimulation and lateral connections between reservoir neurons can be readout by the sink layer neurons. This activity is used for adaptation of synapses between reservoir and sink layer neurons. A new measure called the discriminability index (DI) is introduced to compute if the network can discriminate between old patterns already presented in an initial training session. The DI is also used to compute if the network adapts to new patterns without losing its ability to discriminate among old patterns. The final outcome is that the network is able to correctly discriminate between all patterns-both old and new. This result holds as long as inhibitory synapses employ STDP to continuously enable current balance in the network. The results suggest a possible direction for future investigation into how spiking neural networks could address the stability-plasticity question despite having continuous synaptic plasticity.

Adaptive fuzzy leader clustering of complex data sets in pattern recognition

NASA Technical Reports Server (NTRS)

Newton, Scott C.; Pemmaraju, Surya; Mitra, Sunanda

1992-01-01

A modular, unsupervised neural network architecture for clustering and classification of complex data sets is presented. The adaptive fuzzy leader clustering (AFLC) architecture is a hybrid neural-fuzzy system that learns on-line in a stable and efficient manner. The initial classification is performed in two stages: a simple competitive stage and a distance metric comparison stage. The cluster prototypes are then incrementally updated by relocating the centroid positions from fuzzy C-means system equations for the centroids and the membership values. The AFLC algorithm is applied to the Anderson Iris data and laser-luminescent fingerprint image data. It is concluded that the AFLC algorithm successfully classifies features extracted from real data, discrete or continuous.

[Terahertz Spectroscopic Identification with Deep Belief Network].

PubMed

Ma, Shuai; Shen, Tao; Wang, Rui-qi; Lai, Hua; Yu, Zheng-tao

2015-12-01

Feature extraction and classification are the key issues of terahertz spectroscopy identification. Because many materials have no apparent absorption peaks in the terahertz band, it is difficult to extract theirs terahertz spectroscopy feature and identify. To this end, a novel of identify terahertz spectroscopy approach with Deep Belief Network (DBN) was studied in this paper, which combines the advantages of DBN and K-Nearest Neighbors (KNN) classifier. Firstly, cubic spline interpolation and S-G filter were used to normalize the eight kinds of substances (ATP, Acetylcholine Bromide, Bifenthrin, Buprofezin, Carbazole, Bleomycin, Buckminster and Cylotriphosphazene) terahertz transmission spectra in the range of 0.9-6 THz. Secondly, the DBN model was built by two restricted Boltzmann machine (RBM) and then trained layer by layer using unsupervised approach. Instead of using handmade features, the DBN was employed to learn suitable features automatically with raw input data. Finally, a KNN classifier was applied to identify the terahertz spectrum. Experimental results show that using the feature learned by DBN can identify the terahertz spectrum of different substances with the recognition rate of over 90%, which demonstrates that the proposed method can automatically extract the effective features of terahertz spectrum. Furthermore, this KNN classifier was compared with others (BP neural network, SOM neural network and RBF neural network). Comparisons showed that the recognition rate of KNN classifier is better than the other three classifiers. Using the approach that automatic extract terahertz spectrum features by DBN can greatly reduce the workload of feature extraction. This proposed method shows a promising future in the application of identifying the mass terahertz spectroscopy.

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.

Unsupervised classification of variable stars

NASA Astrophysics Data System (ADS)

Valenzuela, Lucas; Pichara, Karim

2018-03-01

During the past 10 years, a considerable amount of effort has been made to develop algorithms for automatic classification of variable stars. That has been primarily achieved by applying machine learning methods to photometric data sets where objects are represented as light curves. Classifiers require training sets to learn the underlying patterns that allow the separation among classes. Unfortunately, building training sets is an expensive process that demands a lot of human efforts. Every time data come from new surveys; the only available training instances are the ones that have a cross-match with previously labelled objects, consequently generating insufficient training sets compared with the large amounts of unlabelled sources. In this work, we present an algorithm that performs unsupervised classification of variable stars, relying only on the similarity among light curves. We tackle the unsupervised classification problem by proposing an untraditional approach. Instead of trying to match classes of stars with clusters found by a clustering algorithm, we propose a query-based method where astronomers can find groups of variable stars ranked by similarity. We also develop a fast similarity function specific for light curves, based on a novel data structure that allows scaling the search over the entire data set of unlabelled objects. Experiments show that our unsupervised model achieves high accuracy in the classification of different types of variable stars and that the proposed algorithm scales up to massive amounts of light curves.

Central auditory neurons have composite receptive fields.

PubMed

Kozlov, Andrei S; Gentner, Timothy Q

2016-02-02

High-level neurons processing complex, behaviorally relevant signals are sensitive to conjunctions of features. Characterizing the receptive fields of such neurons is difficult with standard statistical tools, however, and the principles governing their organization remain poorly understood. Here, we demonstrate multiple distinct receptive-field features in individual high-level auditory neurons in a songbird, European starling, in response to natural vocal signals (songs). We then show that receptive fields with similar characteristics can be reproduced by an unsupervised neural network trained to represent starling songs with a single learning rule that enforces sparseness and divisive normalization. We conclude that central auditory neurons have composite receptive fields that can arise through a combination of sparseness and normalization in neural circuits. Our results, along with descriptions of random, discontinuous receptive fields in the central olfactory neurons in mammals and insects, suggest general principles of neural computation across sensory systems and animal classes.

SAR image segmentation using skeleton-based fuzzy clustering

NASA Astrophysics Data System (ADS)

Cao, Yun Yi; Chen, Yan Qiu

2003-06-01

SAR image segmentation can be converted to a clustering problem in which pixels or small patches are grouped together based on local feature information. In this paper, we present a novel framework for segmentation. The segmentation goal is achieved by unsupervised clustering upon characteristic descriptors extracted from local patches. The mixture model of characteristic descriptor, which combines intensity and texture feature, is investigated. The unsupervised algorithm is derived from the recently proposed Skeleton-Based Data Labeling method. Skeletons are constructed as prototypes of clusters to represent arbitrary latent structures in image data. Segmentation using Skeleton-Based Fuzzy Clustering is able to detect the types of surfaces appeared in SAR images automatically without any user input.

A Benchmark Dataset and Saliency-guided Stacked Autoencoders for Video-based Salient Object Detection.

PubMed

Li, Jia; Xia, Changqun; Chen, Xiaowu

2017-10-12

Image-based salient object detection (SOD) has been extensively studied in past decades. However, video-based SOD is much less explored due to the lack of large-scale video datasets within which salient objects are unambiguously defined and annotated. Toward this end, this paper proposes a video-based SOD dataset that consists of 200 videos. In constructing the dataset, we manually annotate all objects and regions over 7,650 uniformly sampled keyframes and collect the eye-tracking data of 23 subjects who free-view all videos. From the user data, we find that salient objects in a video can be defined as objects that consistently pop-out throughout the video, and objects with such attributes can be unambiguously annotated by combining manually annotated object/region masks with eye-tracking data of multiple subjects. To the best of our knowledge, it is currently the largest dataset for videobased salient object detection. Based on this dataset, this paper proposes an unsupervised baseline approach for video-based SOD by using saliencyguided stacked autoencoders. In the proposed approach, multiple spatiotemporal saliency cues are first extracted at the pixel, superpixel and object levels. With these saliency cues, stacked autoencoders are constructed in an unsupervised manner that automatically infers a saliency score for each pixel by progressively encoding the high-dimensional saliency cues gathered from the pixel and its spatiotemporal neighbors. In experiments, the proposed unsupervised approach is compared with 31 state-of-the-art models on the proposed dataset and outperforms 30 of them, including 19 imagebased classic (unsupervised or non-deep learning) models, six image-based deep learning models, and five video-based unsupervised models. Moreover, benchmarking results show that the proposed dataset is very challenging and has the potential to boost the development of video-based SOD.

Knowledge discovery by accuracy maximization

PubMed Central

Cacciatore, Stefano; Luchinat, Claudio; Tenori, Leonardo

2014-01-01

Here we describe KODAMA (knowledge discovery by accuracy maximization), an unsupervised and semisupervised learning algorithm that performs feature extraction from noisy and high-dimensional data. Unlike other data mining methods, the peculiarity of KODAMA is that it is driven by an integrated procedure of cross-validation of the results. The discovery of a local manifold’s topology is led by a classifier through a Monte Carlo procedure of maximization of cross-validated predictive accuracy. Briefly, our approach differs from previous methods in that it has an integrated procedure of validation of the results. In this way, the method ensures the highest robustness of the obtained solution. This robustness is demonstrated on experimental datasets of gene expression and metabolomics, where KODAMA compares favorably with other existing feature extraction methods. KODAMA is then applied to an astronomical dataset, revealing unexpected features. Interesting and not easily predictable features are also found in the analysis of the State of the Union speeches by American presidents: KODAMA reveals an abrupt linguistic transition sharply separating all post-Reagan from all pre-Reagan speeches. The transition occurs during Reagan’s presidency and not from its beginning. PMID:24706821

Deep learning algorithms for detecting explosive hazards in ground penetrating radar data

NASA Astrophysics Data System (ADS)

Besaw, Lance E.; Stimac, Philip J.

2014-05-01

Buried explosive hazards (BEHs) have been, and continue to be, one of the most deadly threats in modern conflicts. Current handheld sensors rely on a highly trained operator for them to be effective in detecting BEHs. New algorithms are needed to reduce the burden on the operator and improve the performance of handheld BEH detectors. Traditional anomaly detection and discrimination algorithms use "hand-engineered" feature extraction techniques to characterize and classify threats. In this work we use a Deep Belief Network (DBN) to transcend the traditional approaches of BEH detection (e.g., principal component analysis and real-time novelty detection techniques). DBNs are pretrained using an unsupervised learning algorithm to generate compressed representations of unlabeled input data and form feature detectors. They are then fine-tuned using a supervised learning algorithm to form a predictive model. Using ground penetrating radar (GPR) data collected by a robotic cart swinging a handheld detector, our research demonstrates that relatively small DBNs can learn to model GPR background signals and detect BEHs with an acceptable false alarm rate (FAR). In this work, our DBNs achieved 91% probability of detection (Pd) with 1.4 false alarms per square meter when evaluated on anti-tank and anti-personnel targets at temperate and arid test sites. This research demonstrates that DBNs are a viable approach to detect and classify BEHs.

Neural Evidence of Statistical Learning: Efficient Detection of Visual Regularities without Awareness

ERIC Educational Resources Information Center

Turk-Browne, Nicholas B.; Scholl, Brian J.; Chun, Marvin M.; Johnson, Marcia K.

2009-01-01

Our environment contains regularities distributed in space and time that can be detected by way of statistical learning. This unsupervised learning occurs without intent or awareness, but little is known about how it relates to other types of learning, how it affects perceptual processing, and how quickly it can occur. Here we use fMRI during…

Object-oriented feature-tracking algorithms for SAR images of the marginal ice zone

NASA Technical Reports Server (NTRS)

Daida, Jason; Samadani, Ramin; Vesecky, John F.

1990-01-01

An unsupervised method that chooses and applies the most appropriate tracking algorithm from among different sea-ice tracking algorithms is reported. In contrast to current unsupervised methods, this method chooses and applies an algorithm by partially examining a sequential image pair to draw inferences about what was examined. Based on these inferences the reported method subsequently chooses which algorithm to apply to specific areas of the image pair where that algorithm should work best.

Classification of earth terrain using polarimetric synthetic aperture radar images

NASA Technical Reports Server (NTRS)

Lim, H. H.; Swartz, A. A.; Yueh, H. A.; Kong, J. A.; Shin, R. T.; Van Zyl, J. J.

1989-01-01

Supervised and unsupervised classification techniques are developed and used to classify the earth terrain components from SAR polarimetric images of San Francisco Bay and Traverse City, Michigan. The supervised techniques include the Bayes classifiers, normalized polarimetric classification, and simple feature classification using discriminates such as the absolute and normalized magnitude response of individual receiver channel returns and the phase difference between receiver channels. An algorithm is developed as an unsupervised technique which classifies terrain elements based on the relationship between the orientation angle and the handedness of the transmitting and receiving polariation states. It is found that supervised classification produces the best results when accurate classifier training data are used, while unsupervised classification may be applied when training data are not available.

Some simple guides to finding useful information in exploration geochemical data

USGS Publications Warehouse

Singer, D.A.; Kouda, R.

2001-01-01

Most regional geochemistry data reflect processes that can produce superfluous bits of noise and, perhaps, information about the mineralization process of interest. There are two end-member approaches to finding patterns in geochemical data-unsupervised learning and supervised learning. In unsupervised learning, data are processed and the geochemist is given the task of interpreting and identifying possible sources of any patterns. In supervised learning, data from known subgroups such as rock type, mineralized and nonmineralized, and types of mineralization are used to train the system which then is given unknown samples to classify into these subgroups. To locate patterns of interest, it is helpful to transform the data and to remove unwanted masking patterns. With trace elements use of a logarithmic transformation is recommended. In many situations, missing censored data can be estimated using multiple regression of other uncensored variables on the variable with censored values. In unsupervised learning, transformed values can be standardized, or normalized, to a Z-score by subtracting the subset's mean and dividing by its standard deviation. Subsets include any source of differences that might be related to processes unrelated to the target sought such as different laboratories, regional alteration, analytical procedures, or rock types. Normalization removes effects of different means and measurement scales as well as facilitates comparison of spatial patterns of elements. These adjustments remove effects of different subgroups and hopefully leave on the map the simple and uncluttered pattern(s) related to the mineralization only. Supervised learning methods, such as discriminant analysis and neural networks, offer the promise of consistent and, in certain situations, unbiased estimates of where mineralization might exist. These methods critically rely on being trained with data that encompasses all populations fairly and that can possibly fall into only the identified populations. ?? 2001 International Association for Mathematical Geology.

Automated Interpretation of Subcellular Patterns in Fluorescence Microscope Images for Location Proteomics

PubMed Central

Chen, Xiang; Velliste, Meel; Murphy, Robert F.

2010-01-01

Proteomics, the large scale identification and characterization of many or all proteins expressed in a given cell type, has become a major area of biological research. In addition to information on protein sequence, structure and expression levels, knowledge of a protein’s subcellular location is essential to a complete understanding of its functions. Currently subcellular location patterns are routinely determined by visual inspection of fluorescence microscope images. We review here research aimed at creating systems for automated, systematic determination of location. These employ numerical feature extraction from images, feature reduction to identify the most useful features, and various supervised learning (classification) and unsupervised learning (clustering) methods. These methods have been shown to perform significantly better than human interpretation of the same images. When coupled with technologies for tagging large numbers of proteins and high-throughput microscope systems, the computational methods reviewed here enable the new subfield of location proteomics. This subfield will make critical contributions in two related areas. First, it will provide structured, high-resolution information on location to enable Systems Biology efforts to simulate cell behavior from the gene level on up. Second, it will provide tools for Cytomics projects aimed at characterizing the behaviors of all cell types before, during and after the onset of various diseases. PMID:16752421

Deep learning of joint myelin and T1w MRI features in normal-appearing brain tissue to distinguish between multiple sclerosis patients and healthy controls.

PubMed

Yoo, Youngjin; Tang, Lisa Y W; Brosch, Tom; Li, David K B; Kolind, Shannon; Vavasour, Irene; Rauscher, Alexander; MacKay, Alex L; Traboulsee, Anthony; Tam, Roger C

2018-01-01

Myelin imaging is a form of quantitative magnetic resonance imaging (MRI) that measures myelin content and can potentially allow demyelinating diseases such as multiple sclerosis (MS) to be detected earlier. Although focal lesions are the most visible signs of MS pathology on conventional MRI, it has been shown that even tissues that appear normal may exhibit decreased myelin content as revealed by myelin-specific images (i.e., myelin maps). Current methods for analyzing myelin maps typically use global or regional mean myelin measurements to detect abnormalities, but ignore finer spatial patterns that may be characteristic of MS. In this paper, we present a machine learning method to automatically learn, from multimodal MR images, latent spatial features that can potentially improve the detection of MS pathology at early stage. More specifically, 3D image patches are extracted from myelin maps and the corresponding T1-weighted (T1w) MRIs, and are used to learn a latent joint myelin-T1w feature representation via unsupervised deep learning. Using a data set of images from MS patients and healthy controls, a common set of patches are selected via a voxel-wise t -test performed between the two groups. In each MS image, any patches overlapping with focal lesions are excluded, and a feature imputation method is used to fill in the missing values. A feature selection process (LASSO) is then utilized to construct a sparse representation. The resulting normal-appearing features are used to train a random forest classifier. Using the myelin and T1w images of 55 relapse-remitting MS patients and 44 healthy controls in an 11-fold cross-validation experiment, the proposed method achieved an average classification accuracy of 87.9% (SD = 8.4%), which is higher and more consistent across folds than those attained by regional mean myelin (73.7%, SD = 13.7%) and T1w measurements (66.7%, SD = 10.6%), or deep-learned features in either the myelin (83.8%, SD = 11.0%) or T1w (70.1%, SD = 13.6%) images alone, suggesting that the proposed method has strong potential for identifying image features that are more sensitive and specific to MS pathology in normal-appearing brain tissues.

Unsupervised Scalable Statistical Method for Identifying Influential Users in Online Social Networks.

PubMed

Azcorra, A; Chiroque, L F; Cuevas, R; Fernández Anta, A; Laniado, H; Lillo, R E; Romo, J; Sguera, C

2018-05-03

Billions of users interact intensively every day via Online Social Networks (OSNs) such as Facebook, Twitter, or Google+. This makes OSNs an invaluable source of information, and channel of actuation, for sectors like advertising, marketing, or politics. To get the most of OSNs, analysts need to identify influential users that can be leveraged for promoting products, distributing messages, or improving the image of companies. In this report we propose a new unsupervised method, Massive Unsupervised Outlier Detection (MUOD), based on outliers detection, for providing support in the identification of influential users. MUOD is scalable, and can hence be used in large OSNs. Moreover, it labels the outliers as of shape, magnitude, or amplitude, depending of their features. This allows classifying the outlier users in multiple different classes, which are likely to include different types of influential users. Applying MUOD to a subset of roughly 400 million Google+ users, it has allowed identifying and discriminating automatically sets of outlier users, which present features associated to different definitions of influential users, like capacity to attract engagement, capacity to attract a large number of followers, or high infection capacity.

VizieR Online Data Catalog: Redshift reliability flags (VVDS data) (Jamal+, 2018)

NASA Astrophysics Data System (ADS)

Jamal, S.; Le Brun, V.; Le Fevre, O.; Vibert, D.; Schmitt, A.; Surace, C.; Copin, Y.; Garilli, B.; Moresco, M.; Pozzetti, L.

2017-09-01

The VIMOS VLT Deep Survey (Le Fevre et al. 2013A&A...559A..14L) is a combination of 3 i-band magnitude limited surveys: Wide (17.5<=iAB<=22.5; 8.6deg2), Deep (17.5<=iAB<=24; 0.6deg2) and Ultra-Deep (23<=iAB<=24.75; 512arcmin2), that produced a total of 35526 spectroscopic galaxy redshifts between 0 and 6.7 (22434 in Wide, 12051 in Deep and 1041 in UDeep). We supplement spectra of the VIMOS VLT Deep Survey (VVDS) with newly-defined redshift reliability flags obtained from clustering (unsupervised classification in Machine Learning) a set of descriptors from individual zPDFs. In this paper, we exploit a set of 24519 spectra from the VVDS database. After computing zPDFs for each individual spectrum, a set of (8) descriptors of the zPDF are extracted to build a feature matrix X (dimension = 24519 rows, 8 columns). Then, we use a clustering (unsupervised algorithms in Machine Learning) algorithm to partition the feature space into distinct clusters (5 clusters: C1,C2,C3,C4,C5), each depicting a different level of confidence to associate with the measured redshift zMAP (Maximum-A-Posteriori estimate that corresponds to the maximum of the redshift PDF). The clustering results (C1,C2,C3,C4,C5) reported in the table are those used in the paper (Jamal et al, 2017) to present the new methodology of automating the zspec reliability assessment. In particular, we would like to point out that they were obtained from first tests conducted on the VVDS spectroscopic data (end of 2016). Therefore, the table does not depict immutable results (on-going improvements). Future updates of the VVDS redshift reliability flags can be expected. (1 data file).

Unsupervised deep learning reveals prognostically relevant subtypes of glioblastoma.

PubMed

Young, Jonathan D; Cai, Chunhui; Lu, Xinghua

2017-10-03

One approach to improving the personalized treatment of cancer is to understand the cellular signaling transduction pathways that cause cancer at the level of the individual patient. In this study, we used unsupervised deep learning to learn the hierarchical structure within cancer gene expression data. Deep learning is a group of machine learning algorithms that use multiple layers of hidden units to capture hierarchically related, alternative representations of the input data. We hypothesize that this hierarchical structure learned by deep learning will be related to the cellular signaling system. Robust deep learning model selection identified a network architecture that is biologically plausible. Our model selection results indicated that the 1st hidden layer of our deep learning model should contain about 1300 hidden units to most effectively capture the covariance structure of the input data. This agrees with the estimated number of human transcription factors, which is approximately 1400. This result lends support to our hypothesis that the 1st hidden layer of a deep learning model trained on gene expression data may represent signals related to transcription factor activation. Using the 3rd hidden layer representation of each tumor as learned by our unsupervised deep learning model, we performed consensus clustering on all tumor samples-leading to the discovery of clusters of glioblastoma multiforme with differential survival. One of these clusters contained all of the glioblastoma samples with G-CIMP, a known methylation phenotype driven by the IDH1 mutation and associated with favorable prognosis, suggesting that the hidden units in the 3rd hidden layer representations captured a methylation signal without explicitly using methylation data as input. We also found differentially expressed genes and well-known mutations (NF1, IDH1, EGFR) that were uniquely correlated with each of these clusters. Exploring these unique genes and mutations will allow us to further investigate the disease mechanisms underlying each of these clusters. In summary, we show that a deep learning model can be trained to represent biologically and clinically meaningful abstractions of cancer gene expression data. Understanding what additional relationships these hidden layer abstractions have with the cancer cellular signaling system could have a significant impact on the understanding and treatment of cancer.

Interactive Algorithms for Unsupervised Machine Learning

DTIC Science & Technology

2015-06-01

committee members, Nina Balcan, Sanjoy Dasgupta, and John Langford. Nina’s unbounded energy and her passion for machine learning are qualities that I...52 3.3.2 Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 3.3.3 Real World Experiments...80 4.4.1 Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 4.4.2 Real World

Conduction Delay Learning Model for Unsupervised and Supervised Classification of Spatio-Temporal Spike Patterns

PubMed Central

Matsubara, Takashi

2017-01-01

Precise spike timing is considered to play a fundamental role in communications and signal processing in biological neural networks. Understanding the mechanism of spike timing adjustment would deepen our understanding of biological systems and enable advanced engineering applications such as efficient computational architectures. However, the biological mechanisms that adjust and maintain spike timing remain unclear. Existing algorithms adopt a supervised approach, which adjusts the axonal conduction delay and synaptic efficacy until the spike timings approximate the desired timings. This study proposes a spike timing-dependent learning model that adjusts the axonal conduction delay and synaptic efficacy in both unsupervised and supervised manners. The proposed learning algorithm approximates the Expectation-Maximization algorithm, and classifies the input data encoded into spatio-temporal spike patterns. Even in the supervised classification, the algorithm requires no external spikes indicating the desired spike timings unlike existing algorithms. Furthermore, because the algorithm is consistent with biological models and hypotheses found in existing biological studies, it could capture the mechanism underlying biological delay learning. PMID:29209191

Conduction Delay Learning Model for Unsupervised and Supervised Classification of Spatio-Temporal Spike Patterns.

PubMed

Matsubara, Takashi

2017-01-01

Precise spike timing is considered to play a fundamental role in communications and signal processing in biological neural networks. Understanding the mechanism of spike timing adjustment would deepen our understanding of biological systems and enable advanced engineering applications such as efficient computational architectures. However, the biological mechanisms that adjust and maintain spike timing remain unclear. Existing algorithms adopt a supervised approach, which adjusts the axonal conduction delay and synaptic efficacy until the spike timings approximate the desired timings. This study proposes a spike timing-dependent learning model that adjusts the axonal conduction delay and synaptic efficacy in both unsupervised and supervised manners. The proposed learning algorithm approximates the Expectation-Maximization algorithm, and classifies the input data encoded into spatio-temporal spike patterns. Even in the supervised classification, the algorithm requires no external spikes indicating the desired spike timings unlike existing algorithms. Furthermore, because the algorithm is consistent with biological models and hypotheses found in existing biological studies, it could capture the mechanism underlying biological delay learning.

CHISSL: A Human-Machine Collaboration Space for Unsupervised Learning

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

Arendt, Dustin L.; Komurlu, Caner; Blaha, Leslie M.

We developed CHISSL, a human-machine interface that utilizes supervised machine learning in an unsupervised context to help the user group unlabeled instances by her own mental model. The user primarily interacts via correction (moving a misplaced instance into its correct group) or confirmation (accepting that an instance is placed in its correct group). Concurrent with the user's interactions, CHISSL trains a classification model guided by the user's grouping of the data. It then predicts the group of unlabeled instances and arranges some of these alongside the instances manually organized by the user. We hypothesize that this mode of human andmore » machine collaboration is more effective than Active Learning, wherein the machine decides for itself which instances should be labeled by the user. We found supporting evidence for this hypothesis in a pilot study where we applied CHISSL to organize a collection of handwritten digits.« less

Learning a Generative Probabilistic Grammar of Experience: A Process-Level Model of Language Acquisition

ERIC Educational Resources Information Center

Kolodny, Oren; Lotem, Arnon; Edelman, Shimon

2015-01-01

We introduce a set of biologically and computationally motivated design choices for modeling the learning of language, or of other types of sequential, hierarchically structured experience and behavior, and describe an implemented system that conforms to these choices and is capable of unsupervised learning from raw natural-language corpora. Given…

A comparison of neural network and fuzzy clustering techniques in segmenting magnetic resonance images of the brain

NASA Technical Reports Server (NTRS)

Hall, Lawrence O.; Bensaid, Amine M.; Clarke, Laurence P.; Velthuizen, Robert P.; Silbiger, Martin S.; Bezdek, James C.

1992-01-01

Magnetic resonance (MR) brain section images are segmented and then synthetically colored to give visual representations of the original data with three approaches: the literal and approximate fuzzy c-means unsupervised clustering algorithms and a supervised computational neural network, a dynamic multilayered perception trained with the cascade correlation learning algorithm. Initial clinical results are presented on both normal volunteers and selected patients with brain tumors surrounded by edema. Supervised and unsupervised segmentation techniques provide broadly similar results. Unsupervised fuzzy algorithms were visually observed to show better segmentation when compared with raw image data for volunteer studies. However, for a more complex segmentation problem with tumor/edema or cerebrospinal fluid boundary, where the tissues have similar MR relaxation behavior, inconsistency in rating among experts was observed.

Ensemble learning with trees and rules: supervised, semi-supervised, unsupervised

USDA-ARS?s Scientific Manuscript database

In this article, we propose several new approaches for post processing a large ensemble of conjunctive rules for supervised and semi-supervised learning problems. We show with various examples that for high dimensional regression problems the models constructed by the post processing the rules with ...

An unsupervised video foreground co-localization and segmentation process by incorporating motion cues and frame features

NASA Astrophysics Data System (ADS)

Zhang, Chao; Zhang, Qian; Zheng, Chi; Qiu, Guoping

2018-04-01

Video foreground segmentation is one of the key problems in video processing. In this paper, we proposed a novel and fully unsupervised approach for foreground object co-localization and segmentation of unconstrained videos. We firstly compute both the actual edges and motion boundaries of the video frames, and then align them by their HOG feature maps. Then, by filling the occlusions generated by the aligned edges, we obtained more precise masks about the foreground object. Such motion-based masks could be derived as the motion-based likelihood. Moreover, the color-base likelihood is adopted for the segmentation process. Experimental Results show that our approach outperforms most of the State-of-the-art algorithms.

Domain adaptation via transfer component analysis.

PubMed

Pan, Sinno Jialin; Tsang, Ivor W; Kwok, James T; Yang, Qiang

2011-02-01

Domain adaptation allows knowledge from a source domain to be transferred to a different but related target domain. Intuitively, discovering a good feature representation across domains is crucial. In this paper, we first propose to find such a representation through a new learning method, transfer component analysis (TCA), for domain adaptation. TCA tries to learn some transfer components across domains in a reproducing kernel Hilbert space using maximum mean miscrepancy. In the subspace spanned by these transfer components, data properties are preserved and data distributions in different domains are close to each other. As a result, with the new representations in this subspace, we can apply standard machine learning methods to train classifiers or regression models in the source domain for use in the target domain. Furthermore, in order to uncover the knowledge hidden in the relations between the data labels from the source and target domains, we extend TCA in a semisupervised learning setting, which encodes label information into transfer components learning. We call this extension semisupervised TCA. The main contribution of our work is that we propose a novel dimensionality reduction framework for reducing the distance between domains in a latent space for domain adaptation. We propose both unsupervised and semisupervised feature extraction approaches, which can dramatically reduce the distance between domain distributions by projecting data onto the learned transfer components. Finally, our approach can handle large datasets and naturally lead to out-of-sample generalization. The effectiveness and efficiency of our approach are verified by experiments on five toy datasets and two real-world applications: cross-domain indoor WiFi localization and cross-domain text classification.

Fault Diagnosis Based on Chemical Sensor Data with an Active Deep Neural Network

PubMed Central

Jiang, Peng; Hu, Zhixin; Liu, Jun; Yu, Shanen; Wu, Feng

2016-01-01

Big sensor data provide significant potential for chemical fault diagnosis, which involves the baseline values of security, stability and reliability in chemical processes. A deep neural network (DNN) with novel active learning for inducing chemical fault diagnosis is presented in this study. It is a method using large amount of chemical sensor data, which is a combination of deep learning and active learning criterion to target the difficulty of consecutive fault diagnosis. DNN with deep architectures, instead of shallow ones, could be developed through deep learning to learn a suitable feature representation from raw sensor data in an unsupervised manner using stacked denoising auto-encoder (SDAE) and work through a layer-by-layer successive learning process. The features are added to the top Softmax regression layer to construct the discriminative fault characteristics for diagnosis in a supervised manner. Considering the expensive and time consuming labeling of sensor data in chemical applications, in contrast to the available methods, we employ a novel active learning criterion for the particularity of chemical processes, which is a combination of Best vs. Second Best criterion (BvSB) and a Lowest False Positive criterion (LFP), for further fine-tuning of diagnosis model in an active manner rather than passive manner. That is, we allow models to rank the most informative sensor data to be labeled for updating the DNN parameters during the interaction phase. The effectiveness of the proposed method is validated in two well-known industrial datasets. Results indicate that the proposed method can obtain superior diagnosis accuracy and provide significant performance improvement in accuracy and false positive rate with less labeled chemical sensor data by further active learning compared with existing methods. PMID:27754386

Fault Diagnosis Based on Chemical Sensor Data with an Active Deep Neural Network.

PubMed

Jiang, Peng; Hu, Zhixin; Liu, Jun; Yu, Shanen; Wu, Feng

2016-10-13

Big sensor data provide significant potential for chemical fault diagnosis, which involves the baseline values of security, stability and reliability in chemical processes. A deep neural network (DNN) with novel active learning for inducing chemical fault diagnosis is presented in this study. It is a method using large amount of chemical sensor data, which is a combination of deep learning and active learning criterion to target the difficulty of consecutive fault diagnosis. DNN with deep architectures, instead of shallow ones, could be developed through deep learning to learn a suitable feature representation from raw sensor data in an unsupervised manner using stacked denoising auto-encoder (SDAE) and work through a layer-by-layer successive learning process. The features are added to the top Softmax regression layer to construct the discriminative fault characteristics for diagnosis in a supervised manner. Considering the expensive and time consuming labeling of sensor data in chemical applications, in contrast to the available methods, we employ a novel active learning criterion for the particularity of chemical processes, which is a combination of Best vs. Second Best criterion (BvSB) and a Lowest False Positive criterion (LFP), for further fine-tuning of diagnosis model in an active manner rather than passive manner. That is, we allow models to rank the most informative sensor data to be labeled for updating the DNN parameters during the interaction phase. The effectiveness of the proposed method is validated in two well-known industrial datasets. Results indicate that the proposed method can obtain superior diagnosis accuracy and provide significant performance improvement in accuracy and false positive rate with less labeled chemical sensor data by further active learning compared with existing methods.

An automatic taxonomy of galaxy morphology using unsupervised machine learning

NASA Astrophysics Data System (ADS)

Hocking, Alex; Geach, James E.; Sun, Yi; Davey, Neil

2018-01-01

We present an unsupervised machine learning technique that automatically segments and labels galaxies in astronomical imaging surveys using only pixel data. Distinct from previous unsupervised machine learning approaches used in astronomy we use no pre-selection or pre-filtering of target galaxy type to identify galaxies that are similar. We demonstrate the technique on the Hubble Space Telescope (HST) Frontier Fields. By training the algorithm using galaxies from one field (Abell 2744) and applying the result to another (MACS 0416.1-2403), we show how the algorithm can cleanly separate early and late type galaxies without any form of pre-directed training for what an 'early' or 'late' type galaxy is. We then apply the technique to the HST Cosmic Assembly Near-infrared Deep Extragalactic Legacy Survey (CANDELS) fields, creating a catalogue of approximately 60 000 classifications. We show how the automatic classification groups galaxies of similar morphological (and photometric) type and make the classifications public via a catalogue, a visual catalogue and galaxy similarity search. We compare the CANDELS machine-based classifications to human-classifications from the Galaxy Zoo: CANDELS project. Although there is not a direct mapping between Galaxy Zoo and our hierarchical labelling, we demonstrate a good level of concordance between human and machine classifications. Finally, we show how the technique can be used to identify rarer objects and present lensed galaxy candidates from the CANDELS imaging.

Automatic extraction of road features in urban environments using dense ALS data

NASA Astrophysics Data System (ADS)

Soilán, Mario; Truong-Hong, Linh; Riveiro, Belén; Laefer, Debra

2018-02-01

This paper describes a methodology that automatically extracts semantic information from urban ALS data for urban parameterization and road network definition. First, building façades are segmented from the ground surface by combining knowledge-based information with both voxel and raster data. Next, heuristic rules and unsupervised learning are applied to the ground surface data to distinguish sidewalk and pavement points as a means for curb detection. Then radiometric information was employed for road marking extraction. Using high-density ALS data from Dublin, Ireland, this fully automatic workflow was able to generate a F-score close to 95% for pavement and sidewalk identification with a resolution of 20 cm and better than 80% for road marking detection.

Detection of dominant flow and abnormal events in surveillance video

NASA Astrophysics Data System (ADS)

Kwak, Sooyeong; Byun, Hyeran

2011-02-01

We propose an algorithm for abnormal event detection in surveillance video. The proposed algorithm is based on a semi-unsupervised learning method, a kind of feature-based approach so that it does not detect the moving object individually. The proposed algorithm identifies dominant flow without individual object tracking using a latent Dirichlet allocation model in crowded environments. It can also automatically detect and localize an abnormally moving object in real-life video. The performance tests are taken with several real-life databases, and their results show that the proposed algorithm can efficiently detect abnormally moving objects in real time. The proposed algorithm can be applied to any situation in which abnormal directions or abnormal speeds are detected regardless of direction.

Segmentation of tumor and edema along with healthy tissues of brain using wavelets and neural networks.

PubMed

Demirhan, Ayşe; Toru, Mustafa; Guler, Inan

2015-07-01

Robust brain magnetic resonance (MR) segmentation algorithms are critical to analyze tissues and diagnose tumor and edema in a quantitative way. In this study, we present a new tissue segmentation algorithm that segments brain MR images into tumor, edema, white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF). The detection of the healthy tissues is performed simultaneously with the diseased tissues because examining the change caused by the spread of tumor and edema on healthy tissues is very important for treatment planning. We used T1, T2, and FLAIR MR images of 20 subjects suffering from glial tumor. We developed an algorithm for stripping the skull before the segmentation process. The segmentation is performed using self-organizing map (SOM) that is trained with unsupervised learning algorithm and fine-tuned with learning vector quantization (LVQ). Unlike other studies, we developed an algorithm for clustering the SOM instead of using an additional network. Input feature vector is constructed with the features obtained from stationary wavelet transform (SWT) coefficients. The results showed that average dice similarity indexes are 91% for WM, 87% for GM, 96% for CSF, 61% for tumor, and 77% for edema.

Multilabel user classification using the community structure of online networks

PubMed Central

Papadopoulos, Symeon; Kompatsiaris, Yiannis

2017-01-01

We study the problem of semi-supervised, multi-label user classification of networked data in the online social platform setting. We propose a framework that combines unsupervised community extraction and supervised, community-based feature weighting before training a classifier. We introduce Approximate Regularized Commute-Time Embedding (ARCTE), an algorithm that projects the users of a social graph onto a latent space, but instead of packing the global structure into a matrix of predefined rank, as many spectral and neural representation learning methods do, it extracts local communities for all users in the graph in order to learn a sparse embedding. To this end, we employ an improvement of personalized PageRank algorithms for searching locally in each user’s graph structure. Then, we perform supervised community feature weighting in order to boost the importance of highly predictive communities. We assess our method performance on the problem of user classification by performing an extensive comparative study among various recent methods based on graph embeddings. The comparison shows that ARCTE significantly outperforms the competition in almost all cases, achieving up to 35% relative improvement compared to the second best competing method in terms of F1-score. PMID:28278242

Multilabel user classification using the community structure of online networks.

PubMed

Rizos, Georgios; Papadopoulos, Symeon; Kompatsiaris, Yiannis

2017-01-01

We study the problem of semi-supervised, multi-label user classification of networked data in the online social platform setting. We propose a framework that combines unsupervised community extraction and supervised, community-based feature weighting before training a classifier. We introduce Approximate Regularized Commute-Time Embedding (ARCTE), an algorithm that projects the users of a social graph onto a latent space, but instead of packing the global structure into a matrix of predefined rank, as many spectral and neural representation learning methods do, it extracts local communities for all users in the graph in order to learn a sparse embedding. To this end, we employ an improvement of personalized PageRank algorithms for searching locally in each user's graph structure. Then, we perform supervised community feature weighting in order to boost the importance of highly predictive communities. We assess our method performance on the problem of user classification by performing an extensive comparative study among various recent methods based on graph embeddings. The comparison shows that ARCTE significantly outperforms the competition in almost all cases, achieving up to 35% relative improvement compared to the second best competing method in terms of F1-score.

Hello, Who is Calling?: Can Words Reveal the Social Nature of Conversations?

PubMed

Stark, Anthony; Shafran, Izhak; Kaye, Jeffrey

2012-01-01

This study aims to infer the social nature of conversations from their content automatically. To place this work in context, our motivation stems from the need to understand how social disengagement affects cognitive decline or depression among older adults. For this purpose, we collected a comprehensive and naturalistic corpus comprising of all the incoming and outgoing telephone calls from 10 subjects over the duration of a year. As a first step, we learned a binary classifier to filter out business related conversation, achieving an accuracy of about 85%. This classification task provides a convenient tool to probe the nature of telephone conversations. We evaluated the utility of openings and closing in differentiating personal calls, and find that empirical results on a large corpus do not support the hypotheses by Schegloff and Sacks that personal conversations are marked by unique closing structures. For classifying different types of social relationships such as family vs other, we investigated features related to language use (entropy), hand-crafted dictionary (LIWC) and topics learned using unsupervised latent Dirichlet models (LDA). Our results show that the posteriors over topics from LDA provide consistently higher accuracy (60-81%) compared to LIWC or language use features in distinguishing different types of conversations.

CNN: a speaker recognition system using a cascaded neural network.

PubMed

Zaki, M; Ghalwash, A; Elkouny, A A

1996-05-01

The main emphasis of this paper is to present an approach for combining supervised and unsupervised neural network models to the issue of speaker recognition. To enhance the overall operation and performance of recognition, the proposed strategy integrates the two techniques, forming one global model called the cascaded model. We first present a simple conventional technique based on the distance measured between a test vector and a reference vector for different speakers in the population. This particular distance metric has the property of weighting down the components in those directions along which the intraspeaker variance is large. The reason for presenting this method is to clarify the discrepancy in performance between the conventional and neural network approach. We then introduce the idea of using unsupervised learning technique, presented by the winner-take-all model, as a means of recognition. Due to several tests that have been conducted and in order to enhance the performance of this model, dealing with noisy patterns, we have preceded it with a supervised learning model--the pattern association model--which acts as a filtration stage. This work includes both the design and implementation of both conventional and neural network approaches to recognize the speakers templates--which are introduced to the system via a voice master card and preprocessed before extracting the features used in the recognition. The conclusion indicates that the system performance in case of neural network is better than that of the conventional one, achieving a smooth degradation in respect of noisy patterns, and higher performance in respect of noise-free patterns.

Fluid Lensing based Machine Learning for Augmenting Earth Science Coral Datasets

NASA Astrophysics Data System (ADS)

Li, A.; Instrella, R.; Chirayath, V.

2016-12-01

Recently, there has been increased interest in monitoring the effects of climate change upon the world's marine ecosystems, particularly coral reefs. These delicate ecosystems are especially threatened due to their sensitivity to ocean warming and acidification, leading to unprecedented levels of coral bleaching and die-off in recent years. However, current global aquatic remote sensing datasets are unable to quantify changes in marine ecosystems at spatial and temporal scales relevant to their growth. In this project, we employ various supervised and unsupervised machine learning algorithms to augment existing datasets from NASA's Earth Observing System (EOS), using high resolution airborne imagery. This method utilizes NASA's ongoing airborne campaigns as well as its spaceborne assets to collect remote sensing data over these afflicted regions, and employs Fluid Lensing algorithms to resolve optical distortions caused by the fluid surface, producing cm-scale resolution imagery of these diverse ecosystems from airborne platforms. Support Vector Machines (SVMs) and K-mean clustering methods were applied to satellite imagery at 0.5m resolution, producing segmented maps classifying coral based on percent cover and morphology. Compared to a previous study using multidimensional maximum a posteriori (MAP) estimation to separate these features in high resolution airborne datasets, SVMs are able to achieve above 75% accuracy when augmented with existing MAP estimates, while unsupervised methods such as K-means achieve roughly 68% accuracy, verified by manually segmented reference data provided by a marine biologist. This effort thus has broad applications for coastal remote sensing, by helping marine biologists quantify behavioral trends spanning large areas and over longer timescales, and to assess the health of coral reefs worldwide.

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.

Simultaneous Stimulus Preexposure Enhances Human Tactile Perceptual Learning

ERIC Educational Resources Information Center

Rodríguez, Gabriel; Angulo, Rocío

2014-01-01

An experiment with human participants established a novel procedure to assess perceptual learning with tactile stimuli. Participants received unsupervised exposure to two sandpaper surfaces differing in roughness (A and B). The ability of the participants to discriminate between the stimuli was subsequently assessed on a same/different test. It…

AHaH computing-from metastable switches to attractors to machine learning.

PubMed

Nugent, Michael Alexander; Molter, Timothy Wesley

2014-01-01

Modern computing architecture based on the separation of memory and processing leads to a well known problem called the von Neumann bottleneck, a restrictive limit on the data bandwidth between CPU and RAM. This paper introduces a new approach to computing we call AHaH computing where memory and processing are combined. The idea is based on the attractor dynamics of volatile dissipative electronics inspired by biological systems, presenting an attractive alternative architecture that is able to adapt, self-repair, and learn from interactions with the environment. We envision that both von Neumann and AHaH computing architectures will operate together on the same machine, but that the AHaH computing processor may reduce the power consumption and processing time for certain adaptive learning tasks by orders of magnitude. The paper begins by drawing a connection between the properties of volatility, thermodynamics, and Anti-Hebbian and Hebbian (AHaH) plasticity. We show how AHaH synaptic plasticity leads to attractor states that extract the independent components of applied data streams and how they form a computationally complete set of logic functions. After introducing a general memristive device model based on collections of metastable switches, we show how adaptive synaptic weights can be formed from differential pairs of incremental memristors. We also disclose how arrays of synaptic weights can be used to build a neural node circuit operating AHaH plasticity. By configuring the attractor states of the AHaH node in different ways, high level machine learning functions are demonstrated. This includes unsupervised clustering, supervised and unsupervised classification, complex signal prediction, unsupervised robotic actuation and combinatorial optimization of procedures-all key capabilities of biological nervous systems and modern machine learning algorithms with real world application.

AHaH Computing–From Metastable Switches to Attractors to Machine Learning

PubMed Central

Nugent, Michael Alexander; Molter, Timothy Wesley

2014-01-01

Modern computing architecture based on the separation of memory and processing leads to a well known problem called the von Neumann bottleneck, a restrictive limit on the data bandwidth between CPU and RAM. This paper introduces a new approach to computing we call AHaH computing where memory and processing are combined. The idea is based on the attractor dynamics of volatile dissipative electronics inspired by biological systems, presenting an attractive alternative architecture that is able to adapt, self-repair, and learn from interactions with the environment. We envision that both von Neumann and AHaH computing architectures will operate together on the same machine, but that the AHaH computing processor may reduce the power consumption and processing time for certain adaptive learning tasks by orders of magnitude. The paper begins by drawing a connection between the properties of volatility, thermodynamics, and Anti-Hebbian and Hebbian (AHaH) plasticity. We show how AHaH synaptic plasticity leads to attractor states that extract the independent components of applied data streams and how they form a computationally complete set of logic functions. After introducing a general memristive device model based on collections of metastable switches, we show how adaptive synaptic weights can be formed from differential pairs of incremental memristors. We also disclose how arrays of synaptic weights can be used to build a neural node circuit operating AHaH plasticity. By configuring the attractor states of the AHaH node in different ways, high level machine learning functions are demonstrated. This includes unsupervised clustering, supervised and unsupervised classification, complex signal prediction, unsupervised robotic actuation and combinatorial optimization of procedures–all key capabilities of biological nervous systems and modern machine learning algorithms with real world application. PMID:24520315

Unsupervised feature relevance analysis applied to improve ECG heartbeat clustering.

PubMed

Rodríguez-Sotelo, J L; Peluffo-Ordoñez, D; Cuesta-Frau, D; Castellanos-Domínguez, G

2012-10-01

The computer-assisted analysis of biomedical records has become an essential tool in clinical settings. However, current devices provide a growing amount of data that often exceeds the processing capacity of normal computers. As this amount of information rises, new demands for more efficient data extracting methods appear. This paper addresses the task of data mining in physiological records using a feature selection scheme. An unsupervised method based on relevance analysis is described. This scheme uses a least-squares optimization of the input feature matrix in a single iteration. The output of the algorithm is a feature weighting vector. The performance of the method was assessed using a heartbeat clustering test on real ECG records. The quantitative cluster validity measures yielded a correctly classified heartbeat rate of 98.69% (specificity), 85.88% (sensitivity) and 95.04% (general clustering performance), which is even higher than the performance achieved by other similar ECG clustering studies. The number of features was reduced on average from 100 to 18, and the temporal cost was a 43% lower than in previous ECG clustering schemes. Copyright © 2012 Elsevier Ireland Ltd. All rights reserved.

Teaching children with autism appropriate play in unsupervised environments using a self-management treatment package.

PubMed Central

Stahmer, A C; Schreibman, L

1992-01-01

The present study used a self-management treatment package to teach 3 children with autism, who exhibited inappropriate play behaviors, to play appropriately in the absence of a treatment provider. After self-management training, generalization and maintenance of the behavior change were assessed. Because of the detrimental effects of self-stimulation (arm flapping, spinning toys, twirling, etc.) on learning, the relationship between self-stimulatory behaviors and appropriate play was measured. Results indicated that the children learned to exhibit appropriate play skills in unsupervised settings, appropriate play skills generalized to new settings, and 2 of the children maintained their gains at 1-month follow-up. In addition, self-stimulatory behaviors decreased as appropriate play increased. Treatment implications of these findings are discussed. PMID:1634432

Mastication Evaluation With Unsupervised Learning: Using an Inertial Sensor-Based System

PubMed Central

Lucena, Caroline Vieira; Lacerda, Marcelo; Caldas, Rafael; De Lima Neto, Fernando Buarque

2018-01-01

There is a direct relationship between the prevalence of musculoskeletal disorders of the temporomandibular joint and orofacial disorders. A well-elaborated analysis of the jaw movements provides relevant information for healthcare professionals to conclude their diagnosis. Different approaches have been explored to track jaw movements such that the mastication analysis is getting less subjective; however, all methods are still highly subjective, and the quality of the assessments depends much on the experience of the health professional. In this paper, an accurate and non-invasive method based on a commercial low-cost inertial sensor (MPU6050) to measure jaw movements is proposed. The jaw-movement feature values are compared to the obtained with clinical analysis, showing no statistically significant difference between both methods. Moreover, We propose to use unsupervised paradigm approaches to cluster mastication patterns of healthy subjects and simulated patients with facial trauma. Two techniques were used in this paper to instantiate the method: Kohonen’s Self-Organizing Maps and K-Means Clustering. Both algorithms have excellent performances to process jaw-movements data, showing encouraging results and potential to bring a full assessment of the masticatory function. The proposed method can be applied in real-time providing relevant dynamic information for health-care professionals. PMID:29651365

Do deep convolutional neural networks really need to be deep when applied for remote scene classification?

NASA Astrophysics Data System (ADS)

Luo, Chang; Wang, Jie; Feng, Gang; Xu, Suhui; Wang, Shiqiang

2017-10-01

Deep convolutional neural networks (CNNs) have been widely used to obtain high-level representation in various computer vision tasks. However, for remote scene classification, there are not sufficient images to train a very deep CNN from scratch. From two viewpoints of generalization power, we propose two promising kinds of deep CNNs for remote scenes and try to find whether deep CNNs need to be deep for remote scene classification. First, we transfer successful pretrained deep CNNs to remote scenes based on the theory that depth of CNNs brings the generalization power by learning available hypothesis for finite data samples. Second, according to the opposite viewpoint that generalization power of deep CNNs comes from massive memorization and shallow CNNs with enough neural nodes have perfect finite sample expressivity, we design a lightweight deep CNN (LDCNN) for remote scene classification. With five well-known pretrained deep CNNs, experimental results on two independent remote-sensing datasets demonstrate that transferred deep CNNs can achieve state-of-the-art results in an unsupervised setting. However, because of its shallow architecture, LDCNN cannot obtain satisfactory performance, regardless of whether in an unsupervised, semisupervised, or supervised setting. CNNs really need depth to obtain general features for remote scenes. This paper also provides baseline for applying deep CNNs to other remote sensing tasks.

Classification of ROTSE Variable Stars using Machine Learning

NASA Astrophysics Data System (ADS)

Wozniak, P. R.; Akerlof, C.; Amrose, S.; Brumby, S.; Casperson, D.; Gisler, G.; Kehoe, R.; Lee, B.; Marshall, S.; McGowan, K. E.; McKay, T.; Perkins, S.; Priedhorsky, W.; Rykoff, E.; Smith, D. A.; Theiler, J.; Vestrand, W. T.; Wren, J.; ROTSE Collaboration

2001-12-01

We evaluate several Machine Learning algorithms as potential tools for automated classification of variable stars. Using the ROTSE sample of ~1800 variables from a pilot study of 5% of the whole sky, we compare the effectiveness of a supervised technique (Support Vector Machines, SVM) versus unsupervised methods (K-means and Autoclass). There are 8 types of variables in the sample: RR Lyr AB, RR Lyr C, Delta Scuti, Cepheids, detached eclipsing binaries, contact binaries, Miras and LPVs. Preliminary results suggest a very high ( ~95%) efficiency of SVM in isolating a few best defined classes against the rest of the sample, and good accuracy ( ~70-75%) for all classes considered simultaneously. This includes some degeneracies, irreducible with the information at hand. Supervised methods naturally outperform unsupervised methods, in terms of final error rate, but unsupervised methods offer many advantages for large sets of unlabeled data. Therefore, both types of methods should be considered as promising tools for mining vast variability surveys. We project that there are more than 30,000 periodic variables in the ROTSE-I data base covering the entire local sky between V=10 and 15.5 mag. This sample size is already stretching the time capabilities of human analysts.

Integrative analysis of gene expression and DNA methylation using unsupervised feature extraction for detecting candidate cancer biomarkers.

PubMed

Moon, Myungjin; Nakai, Kenta

2018-04-01

Currently, cancer biomarker discovery is one of the important research topics worldwide. In particular, detecting significant genes related to cancer is an important task for early diagnosis and treatment of cancer. Conventional studies mostly focus on genes that are differentially expressed in different states of cancer; however, noise in gene expression datasets and insufficient information in limited datasets impede precise analysis of novel candidate biomarkers. In this study, we propose an integrative analysis of gene expression and DNA methylation using normalization and unsupervised feature extractions to identify candidate biomarkers of cancer using renal cell carcinoma RNA-seq datasets. Gene expression and DNA methylation datasets are normalized by Box-Cox transformation and integrated into a one-dimensional dataset that retains the major characteristics of the original datasets by unsupervised feature extraction methods, and differentially expressed genes are selected from the integrated dataset. Use of the integrated dataset demonstrated improved performance as compared with conventional approaches that utilize gene expression or DNA methylation datasets alone. Validation based on the literature showed that a considerable number of top-ranked genes from the integrated dataset have known relationships with cancer, implying that novel candidate biomarkers can also be acquired from the proposed analysis method. Furthermore, we expect that the proposed method can be expanded for applications involving various types of multi-omics datasets.

Identification of More Feasible MicroRNA-mRNA Interactions within Multiple Cancers Using Principal Component Analysis Based Unsupervised Feature Extraction.

PubMed

Taguchi, Y-H

2016-05-10

MicroRNA(miRNA)-mRNA interactions are important for understanding many biological processes, including development, differentiation and disease progression, but their identification is highly context-dependent. When computationally derived from sequence information alone, the identification should be verified by integrated analyses of mRNA and miRNA expression. The drawback of this strategy is the vast number of identified interactions, which prevents an experimental or detailed investigation of each pair. In this paper, we overcome this difficulty by the recently proposed principal component analysis (PCA)-based unsupervised feature extraction (FE), which reduces the number of identified miRNA-mRNA interactions that properly discriminate between patients and healthy controls without losing biological feasibility. The approach is applied to six cancers: hepatocellular carcinoma, non-small cell lung cancer, esophageal squamous cell carcinoma, prostate cancer, colorectal/colon cancer and breast cancer. In PCA-based unsupervised FE, the significance does not depend on the number of samples (as in the standard case) but on the number of features, which approximates the number of miRNAs/mRNAs. To our knowledge, we have newly identified miRNA-mRNA interactions in multiple cancers based on a single common (universal) criterion. Moreover, the number of identified interactions was sufficiently small to be sequentially curated by literature searches.

A new bidirectional heteroassociative memory encompassing correlational, competitive and topological properties.

PubMed

Chartier, Sylvain; Giguère, Gyslain; Langlois, Dominic

2009-01-01

In this paper, we present a new recurrent bidirectional model that encompasses correlational, competitive and topological model properties. The simultaneous use of many classes of network behaviors allows for the unsupervised learning/categorization of perceptual patterns (through input compression) and the concurrent encoding of proximities in a multidimensional space. All of these operations are achieved within a common learning operation, and using a single set of defining properties. It is shown that the model can learn categories by developing prototype representations strictly from exposition to specific exemplars. Moreover, because the model is recurrent, it can reconstruct perfect outputs from incomplete and noisy patterns. Empirical exploration of the model's properties and performance shows that its ability for adequate clustering stems from: (1) properly distributing connection weights, and (2) producing a weight space with a low dispersion level (or higher density). In addition, since the model uses a sparse representation (k-winners), the size of topological neighborhood can be fixed, and no longer requires a decrease through time as was the case with classic self-organizing feature maps. Since the model's learning and transmission parameters are independent from learning trials, the model can develop stable fixed points in a constrained topological architecture, while being flexible enough to learn novel patterns.

Exploring Genome-Wide Expression Profiles Using Machine Learning Techniques.

PubMed

Kebschull, Moritz; Papapanou, Panos N

2017-01-01

Although contemporary high-throughput -omics methods produce high-dimensional data, the resulting wealth of information is difficult to assess using traditional statistical procedures. Machine learning methods facilitate the detection of additional patterns, beyond the mere identification of lists of features that differ between groups.Here, we demonstrate the utility of (1) supervised classification algorithms in class validation, and (2) unsupervised clustering in class discovery. We use data from our previous work that described the transcriptional profiles of gingival tissue samples obtained from subjects suffering from chronic or aggressive periodontitis (1) to test whether the two diagnostic entities were also characterized by differences on the molecular level, and (2) to search for a novel, alternative classification of periodontitis based on the tissue transcriptomes.Using machine learning technology, we provide evidence for diagnostic imprecision in the currently accepted classification of periodontitis, and demonstrate that a novel, alternative classification based on differences in gingival tissue transcriptomes is feasible. The outlined procedures allow for the unbiased interrogation of high-dimensional datasets for characteristic underlying classes, and are applicable to a broad range of -omics data.

Generalized Self-Organizing Maps for Automatic Determination of the Number of Clusters and Their Multiprototypes in Cluster Analysis.

PubMed

Gorzalczany, Marian B; Rudzinski, Filip

2017-06-07

This paper presents a generalization of self-organizing maps with 1-D neighborhoods (neuron chains) that can be effectively applied to complex cluster analysis problems. The essence of the generalization consists in introducing mechanisms that allow the neuron chain--during learning--to disconnect into subchains, to reconnect some of the subchains again, and to dynamically regulate the overall number of neurons in the system. These features enable the network--working in a fully unsupervised way (i.e., using unlabeled data without a predefined number of clusters)--to automatically generate collections of multiprototypes that are able to represent a broad range of clusters in data sets. First, the operation of the proposed approach is illustrated on some synthetic data sets. Then, this technique is tested using several real-life, complex, and multidimensional benchmark data sets available from the University of California at Irvine (UCI) Machine Learning repository and the Knowledge Extraction based on Evolutionary Learning data set repository. A sensitivity analysis of our approach to changes in control parameters and a comparative analysis with an alternative approach are also performed.

Using optimization models to demonstrate the need for structural changes in training programs for surgical medical residents.

PubMed

Turner, Jonathan; Kim, Kibaek; Mehrotra, Sanjay; DaRosa, Debra A; Daskin, Mark S; Rodriguez, Heron E

2013-09-01

The primary goal of a residency program is to prepare trainees for unsupervised care. Duty hour restrictions imposed throughout the prior decade require that residents work significantly fewer hours. Moreover, various stakeholders (e.g. the hospital, mentors, other residents, educators, and patients) require them to prioritize very different activities, often conflicting with their learning goals. Surgical residents' learning goals include providing continuity throughout a patient's pre-, peri-, and post-operative care as well as achieving sufficient surgical experience levels in various procedure types and participating in various formal educational activities, among other things. To complicate matters, senior residents often compete with other residents for surgical experience. This paper features experiments using an optimization model and a real dataset. The experiments test the viability of achieving the above goals at a major academic center using existing models of delivering medical education and training to surgical residents. It develops a detailed multi-objective, two-stage stochastic optimization model with anticipatory capabilities solved over a rolling time horizon. A novel feature of the models is the incorporation of learning curve theory in the objection function. Using a deterministic version of the model, we identify bounds on the achievement of learning goals under existing training paradigms. The computational results highlight the structural problems in the current surgical resident educational system. These results further corroborate earlier findings and suggest an educational system redesign is necessary for surgical medical residents.

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

PubMed

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

2017-01-01

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

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

PubMed Central

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

2017-01-01

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

SUSTAIN: A Network Model of Category Learning

ERIC Educational Resources Information Center

Love, Bradley C.; Medin, Douglas L.; Gureckis, Todd M.

2004-01-01

SUSTAIN (Supervised and Unsupervised STratified Adaptive Incremental Network) is a model of how humans learn categories from examples. SUSTAIN initially assumes a simple category structure. If simple solutions prove inadequate and SUSTAIN is confronted with a surprising event (e.g., it is told that a bat is a mammal instead of a bird), SUSTAIN…

Noise-enhanced clustering and competitive learning algorithms.

PubMed

Osoba, Osonde; Kosko, Bart

2013-01-01

Noise can provably speed up convergence in many centroid-based clustering algorithms. This includes the popular k-means clustering algorithm. The clustering noise benefit follows from the general noise benefit for the expectation-maximization algorithm because many clustering algorithms are special cases of the expectation-maximization algorithm. Simulations show that noise also speeds up convergence in stochastic unsupervised competitive learning, supervised competitive learning, and differential competitive learning. Copyright © 2012 Elsevier Ltd. All rights reserved.

Time-Frequency Learning Machines for Nonstationarity Detection Using Surrogates

NASA Astrophysics Data System (ADS)

Borgnat, Pierre; Flandrin, Patrick; Richard, Cédric; Ferrari, André; Amoud, Hassan; Honeine, Paul

2012-03-01

Time-frequency representations provide a powerful tool for nonstationary signal analysis and classification, supporting a wide range of applications [12]. As opposed to conventional Fourier analysis, these techniques reveal the evolution in time of the spectral content of signals. In Ref. [7,38], time-frequency analysis is used to test stationarity of any signal. The proposed method consists of a comparison between global and local time-frequency features. The originality is to make use of a family of stationary surrogate signals for defining the null hypothesis of stationarity and, based upon this information, to derive statistical tests. An open question remains, however, about how to choose relevant time-frequency features. Over the last decade, a number of new pattern recognition methods based on reproducing kernels have been introduced. These learning machines have gained popularity due to their conceptual simplicity and their outstanding performance [30]. Initiated by Vapnik’s support vector machines (SVM) [35], they offer now a wide class of supervised and unsupervised learning algorithms. In Ref. [17-19], the authors have shown how the most effective and innovative learning machines can be tuned to operate in the time-frequency domain. This chapter follows this line of research by taking advantage of learning machines to test and quantify stationarity. Based on one-class SVM, our approach uses the entire time-frequency representation and does not require arbitrary feature extraction. Applied to a set of surrogates, it provides the domain boundary that includes most of these stationarized signals. This allows us to test the stationarity of the signal under investigation. This chapter is organized as follows. In Section 22.2, we introduce the surrogate data method to generate stationarized signals, namely, the null hypothesis of stationarity. The concept of time-frequency learning machines is presented in Section 22.3, and applied to one-class SVM in order to derive a stationarity test in Section 22.4. The relevance of the latter is illustrated by simulation results in Section 22.5.

Unintentional Drowning

MedlinePlus

... area unsupervised. If you are in and around natural water settings: Use U.S. Coast Guard approved life ... Swimming Pools CDC Feature Article: Drowning Risks in Natural Water Settings CDC: Recreational Water Illnesses (RWIs) CDC ...

The Convallis Rule for Unsupervised Learning in Cortical Networks

PubMed Central

Yger, Pierre; Harris, Kenneth D.

2013-01-01

The phenomenology and cellular mechanisms of cortical synaptic plasticity are becoming known in increasing detail, but the computational principles by which cortical plasticity enables the development of sensory representations are unclear. Here we describe a framework for cortical synaptic plasticity termed the “Convallis rule”, mathematically derived from a principle of unsupervised learning via constrained optimization. Implementation of the rule caused a recurrent cortex-like network of simulated spiking neurons to develop rate representations of real-world speech stimuli, enabling classification by a downstream linear decoder. Applied to spike patterns used in in vitro plasticity experiments, the rule reproduced multiple results including and beyond STDP. However STDP alone produced poorer learning performance. The mathematical form of the rule is consistent with a dual coincidence detector mechanism that has been suggested by experiments in several synaptic classes of juvenile neocortex. Based on this confluence of normative, phenomenological, and mechanistic evidence, we suggest that the rule may approximate a fundamental computational principle of the neocortex. PMID:24204224

Geological applications of machine learning on hyperspectral remote sensing data

NASA Astrophysics Data System (ADS)

Tse, C. H.; Li, Yi-liang; Lam, Edmund Y.

2015-02-01

The CRISM imaging spectrometer orbiting Mars has been producing a vast amount of data in the visible to infrared wavelengths in the form of hyperspectral data cubes. These data, compared with those obtained from previous remote sensing techniques, yield an unprecedented level of detailed spectral resolution in additional to an ever increasing level of spatial information. A major challenge brought about by the data is the burden of processing and interpreting these datasets and extract the relevant information from it. This research aims at approaching the challenge by exploring machine learning methods especially unsupervised learning to achieve cluster density estimation and classification, and ultimately devising an efficient means leading to identification of minerals. A set of software tools have been constructed by Python to access and experiment with CRISM hyperspectral cubes selected from two specific Mars locations. A machine learning pipeline is proposed and unsupervised learning methods were implemented onto pre-processed datasets. The resulting data clusters are compared with the published ASTER spectral library and browse data products from the Planetary Data System (PDS). The result demonstrated that this approach is capable of processing the huge amount of hyperspectral data and potentially providing guidance to scientists for more detailed studies.

Robust Head-Pose Estimation Based on Partially-Latent Mixture of Linear Regressions.

PubMed

Drouard, Vincent; Horaud, Radu; Deleforge, Antoine; Ba, Sileye; Evangelidis, Georgios

2017-03-01

Head-pose estimation has many applications, such as social event analysis, human-robot and human-computer interaction, driving assistance, and so forth. Head-pose estimation is challenging, because it must cope with changing illumination conditions, variabilities in face orientation and in appearance, partial occlusions of facial landmarks, as well as bounding-box-to-face alignment errors. We propose to use a mixture of linear regressions with partially-latent output. This regression method learns to map high-dimensional feature vectors (extracted from bounding boxes of faces) onto the joint space of head-pose angles and bounding-box shifts, such that they are robustly predicted in the presence of unobservable phenomena. We describe in detail the mapping method that combines the merits of unsupervised manifold learning techniques and of mixtures of regressions. We validate our method with three publicly available data sets and we thoroughly benchmark four variants of the proposed algorithm with several state-of-the-art head-pose estimation methods.

Unsupervised detection and removal of muscle artifacts from scalp EEG recordings using canonical correlation analysis, wavelets and random forests.

PubMed

Anastasiadou, Maria N; Christodoulakis, Manolis; Papathanasiou, Eleftherios S; Papacostas, Savvas S; Mitsis, Georgios D

2017-09-01

This paper proposes supervised and unsupervised algorithms for automatic muscle artifact detection and removal from long-term EEG recordings, which combine canonical correlation analysis (CCA) and wavelets with random forests (RF). The proposed algorithms first perform CCA and continuous wavelet transform of the canonical components to generate a number of features which include component autocorrelation values and wavelet coefficient magnitude values. A subset of the most important features is subsequently selected using RF and labelled observations (supervised case) or synthetic data constructed from the original observations (unsupervised case). The proposed algorithms are evaluated using realistic simulation data as well as 30min epochs of non-invasive EEG recordings obtained from ten patients with epilepsy. We assessed the performance of the proposed algorithms using classification performance and goodness-of-fit values for noisy and noise-free signal windows. In the simulation study, where the ground truth was known, the proposed algorithms yielded almost perfect performance. In the case of experimental data, where expert marking was performed, the results suggest that both the supervised and unsupervised algorithm versions were able to remove artifacts without affecting noise-free channels considerably, outperforming standard CCA, independent component analysis (ICA) and Lagged Auto-Mutual Information Clustering (LAMIC). The proposed algorithms achieved excellent performance for both simulation and experimental data. Importantly, for the first time to our knowledge, we were able to perform entirely unsupervised artifact removal, i.e. without using already marked noisy data segments, achieving performance that is comparable to the supervised case. Overall, the results suggest that the proposed algorithms yield significant future potential for improving EEG signal quality in research or clinical settings without the need for marking by expert neurophysiologists, EMG signal recording and user visual inspection. Copyright © 2017 International Federation of Clinical Neurophysiology. Published by Elsevier B.V. All rights reserved.

Adaptive Water Sampling based on Unsupervised Clustering

NASA Astrophysics Data System (ADS)

Py, F.; Ryan, J.; Rajan, K.; Sherman, A.; Bird, L.; Fox, M.; Long, D.

2007-12-01

Autonomous Underwater Vehicles (AUVs) are widely used for oceanographic surveys, during which data is collected from a number of on-board sensors. Engineers and scientists at MBARI have extended this approach by developing a water sampler specialy for the AUV, which can sample a specific patch of water at a specific time. The sampler, named the Gulper, captures 2 liters of seawater in less than 2 seconds on a 21" MBARI Odyssey AUV. Each sample chamber of the Gulper is filled with seawater through a one-way valve, which protrudes through the fairing of the AUV. This new kind of device raises a new problem: when to trigger the gulper autonomously? For example, scientists interested in studying the mobilization and transport of shelf sediments would like to detect intermediate nepheloïd layers (INLs). To be able to detect this phenomenon we need to extract a model based on AUV sensors that can detect this feature in-situ. The formation of such a model is not obvious as identification of this feature is generally based on data from multiple sensors. We have developed an unsupervised data clustering technique to extract the different features which will then be used for on-board classification and triggering of the Gulper. We use a three phase approach: 1) use data from past missions to learn the different classes of data from sensor inputs. The clustering algorithm will then extract the set of features that can be distinguished within this large data set. 2) Scientists on shore then identify these features and point out which correspond to those of interest (e.g. nepheloïd layer, upwelling material etc) 3) Embed the corresponding classifier into the AUV control system to indicate the most probable feature of the water depending on sensory input. The triggering algorithm looks to this result and triggers the Gulper if the classifier indicates that we are within the feature of interest with a predetermined threshold of confidence. We have deployed this method of online classification and sampling based on AUV depth and HOBI Labs Hydroscat-2 sensor data. Using approximately 20,000 data samples the clustering algorithm generated 14 clusters with one identified as corresponding to a nepheloïd layer. We demonstrate that such a technique can be used to reliably and efficiently sample water based on multiple sources of data in real-time.

PepArML: A Meta-Search Peptide Identification Platform

PubMed Central

Edwards, Nathan J.

2014-01-01

The PepArML meta-search peptide identification platform provides a unified search interface to seven search engines; a robust cluster, grid, and cloud computing scheduler for large-scale searches; and an unsupervised, model-free, machine-learning-based result combiner, which selects the best peptide identification for each spectrum, estimates false-discovery rates, and outputs pepXML format identifications. The meta-search platform supports Mascot; Tandem with native, k-score, and s-score scoring; OMSSA; MyriMatch; and InsPecT with MS-GF spectral probability scores — reformatting spectral data and constructing search configurations for each search engine on the fly. The combiner selects the best peptide identification for each spectrum based on search engine results and features that model enzymatic digestion, retention time, precursor isotope clusters, mass accuracy, and proteotypic peptide properties, requiring no prior knowledge of feature utility or weighting. The PepArML meta-search peptide identification platform often identifies 2–3 times more spectra than individual search engines at 10% FDR. PMID:25663956

Accumulating pyramid spatial-spectral collaborative coding divergence for hyperspectral anomaly detection

NASA Astrophysics Data System (ADS)

Sun, Hao; Zou, Huanxin; Zhou, Shilin

2016-03-01

Detection of anomalous targets of various sizes in hyperspectral data has received a lot of attention in reconnaissance and surveillance applications. Many anomaly detectors have been proposed in literature. However, current methods are susceptible to anomalies in the processing window range and often make critical assumptions about the distribution of the background data. Motivated by the fact that anomaly pixels are often distinctive from their local background, in this letter, we proposed a novel hyperspectral anomaly detection framework for real-time remote sensing applications. The proposed framework consists of four major components, sparse feature learning, pyramid grid window selection, joint spatial-spectral collaborative coding and multi-level divergence fusion. It exploits the collaborative representation difference in the feature space to locate potential anomalies and is totally unsupervised without any prior assumptions. Experimental results on airborne recorded hyperspectral data demonstrate that the proposed methods adaptive to anomalies in a large range of sizes and is well suited for parallel processing.

Classifying seismic noise and sources from OBS data using unsupervised machine learning

NASA Astrophysics Data System (ADS)

Mosher, S. G.; Audet, P.

2017-12-01

The paradigm of plate tectonics was established mainly by recognizing the central role of oceanic plates in the production and destruction of tectonic plates at their boundaries. Since that realization, however, seismic studies of tectonic plates and their associated deformation have slowly shifted their attention toward continental plates due to the ease of installation and maintenance of high-quality seismic networks on land. The result has been a much more detailed understanding of the seismicity patterns associated with continental plate deformation in comparison with the low-magnitude deformation patterns within oceanic plates and at their boundaries. While the number of high-quality ocean-bottom seismometer (OBS) deployments within the past decade has demonstrated the potential to significantly increase our understanding of tectonic systems in oceanic settings, OBS data poses significant challenges to many of the traditional data processing techniques in seismology. In particular, problems involving the detection, location, and classification of seismic sources occurring within oceanic settings are much more difficult due to the extremely noisy seafloor environment in which data are recorded. However, classifying data without a priori constraints is a problem that is routinely pursued via unsupervised machine learning algorithms, which remain robust even in cases involving complicated datasets. In this research, we apply simple unsupervised machine learning algorithms (e.g., clustering) to OBS data from the Cascadia Initiative in an attempt to classify and detect a broad range of seismic sources, including various noise sources and tremor signals occurring within ocean settings.

Quasi-Supervised Scoring of Human Sleep in Polysomnograms Using Augmented Input Variables

PubMed Central

Yaghouby, Farid; Sunderam, Sridhar

2015-01-01

The limitations of manual sleep scoring make computerized methods highly desirable. Scoring errors can arise from human rater uncertainty or inter-rater variability. Sleep scoring algorithms either come as supervised classifiers that need scored samples of each state to be trained, or as unsupervised classifiers that use heuristics or structural clues in unscored data to define states. We propose a quasi-supervised classifier that models observations in an unsupervised manner but mimics a human rater wherever training scores are available. EEG, EMG, and EOG features were extracted in 30s epochs from human-scored polysomnograms recorded from 42 healthy human subjects (18 to 79 years) and archived in an anonymized, publicly accessible database. Hypnograms were modified so that: 1. Some states are scored but not others; 2. Samples of all states are scored but not for transitional epochs; and 3. Two raters with 67% agreement are simulated. A framework for quasi-supervised classification was devised in which unsupervised statistical models—specifically Gaussian mixtures and hidden Markov models—are estimated from unlabeled training data, but the training samples are augmented with variables whose values depend on available scores. Classifiers were fitted to signal features incorporating partial scores, and used to predict scores for complete recordings. Performance was assessed using Cohen's K statistic. The quasi-supervised classifier performed significantly better than an unsupervised model and sometimes as well as a completely supervised model despite receiving only partial scores. The quasi-supervised algorithm addresses the need for classifiers that mimic scoring patterns of human raters while compensating for their limitations. PMID:25679475

Quasi-supervised scoring of human sleep in polysomnograms using augmented input variables.

PubMed

Yaghouby, Farid; Sunderam, Sridhar

2015-04-01

The limitations of manual sleep scoring make computerized methods highly desirable. Scoring errors can arise from human rater uncertainty or inter-rater variability. Sleep scoring algorithms either come as supervised classifiers that need scored samples of each state to be trained, or as unsupervised classifiers that use heuristics or structural clues in unscored data to define states. We propose a quasi-supervised classifier that models observations in an unsupervised manner but mimics a human rater wherever training scores are available. EEG, EMG, and EOG features were extracted in 30s epochs from human-scored polysomnograms recorded from 42 healthy human subjects (18-79 years) and archived in an anonymized, publicly accessible database. Hypnograms were modified so that: 1. Some states are scored but not others; 2. Samples of all states are scored but not for transitional epochs; and 3. Two raters with 67% agreement are simulated. A framework for quasi-supervised classification was devised in which unsupervised statistical models-specifically Gaussian mixtures and hidden Markov models--are estimated from unlabeled training data, but the training samples are augmented with variables whose values depend on available scores. Classifiers were fitted to signal features incorporating partial scores, and used to predict scores for complete recordings. Performance was assessed using Cohen's Κ statistic. The quasi-supervised classifier performed significantly better than an unsupervised model and sometimes as well as a completely supervised model despite receiving only partial scores. The quasi-supervised algorithm addresses the need for classifiers that mimic scoring patterns of human raters while compensating for their limitations. Copyright © 2015 Elsevier Ltd. All rights reserved.

Hyperparameterization of soil moisture statistical models for North America with Ensemble Learning Models (Elm)

NASA Astrophysics Data System (ADS)

Steinberg, P. D.; Brener, G.; Duffy, D.; Nearing, G. S.; Pelissier, C.

2017-12-01

Hyperparameterization, of statistical models, i.e. automated model scoring and selection, such as evolutionary algorithms, grid searches, and randomized searches, can improve forecast model skill by reducing errors associated with model parameterization, model structure, and statistical properties of training data. Ensemble Learning Models (Elm), and the related Earthio package, provide a flexible interface for automating the selection of parameters and model structure for machine learning models common in climate science and land cover classification, offering convenient tools for loading NetCDF, HDF, Grib, or GeoTiff files, decomposition methods like PCA and manifold learning, and parallel training and prediction with unsupervised and supervised classification, clustering, and regression estimators. Continuum Analytics is using Elm to experiment with statistical soil moisture forecasting based on meteorological forcing data from NASA's North American Land Data Assimilation System (NLDAS). There Elm is using the NSGA-2 multiobjective optimization algorithm for optimizing statistical preprocessing of forcing data to improve goodness-of-fit for statistical models (i.e. feature engineering). This presentation will discuss Elm and its components, including dask (distributed task scheduling), xarray (data structures for n-dimensional arrays), and scikit-learn (statistical preprocessing, clustering, classification, regression), and it will show how NSGA-2 is being used for automate selection of soil moisture forecast statistical models for North America.

Learning of Chunking Sequences in Cognition and Behavior

PubMed Central

Rabinovich, Mikhail

2015-01-01

We often learn and recall long sequences in smaller segments, such as a phone number 858 534 22 30 memorized as four segments. Behavioral experiments suggest that humans and some animals employ this strategy of breaking down cognitive or behavioral sequences into chunks in a wide variety of tasks, but the dynamical principles of how this is achieved remains unknown. Here, we study the temporal dynamics of chunking for learning cognitive sequences in a chunking representation using a dynamical model of competing modes arranged to evoke hierarchical Winnerless Competition (WLC) dynamics. Sequential memory is represented as trajectories along a chain of metastable fixed points at each level of the hierarchy, and bistable Hebbian dynamics enables the learning of such trajectories in an unsupervised fashion. Using computer simulations, we demonstrate the learning of a chunking representation of sequences and their robust recall. During learning, the dynamics associates a set of modes to each information-carrying item in the sequence and encodes their relative order. During recall, hierarchical WLC guarantees the robustness of the sequence order when the sequence is not too long. The resulting patterns of activities share several features observed in behavioral experiments, such as the pauses between boundaries of chunks, their size and their duration. Failures in learning chunking sequences provide new insights into the dynamical causes of neurological disorders such as Parkinson’s disease and Schizophrenia. PMID:26584306

Learning LM Specificity for Ganglion Cells

NASA Technical Reports Server (NTRS)

Ahumada, Albert J.

2015-01-01

Unsupervised learning models have been proposed based on experience (Ahumada and Mulligan, 1990;Wachtler, Doi, Lee and Sejnowski, 2007) that allow the cortex to develop units with LM specific color opponent receptive fields like the blob cells reported by Hubel and Wiesel on the basis of visual experience. These models used ganglion cells with LM indiscriminate wiring as inputs to the learning mechanism, which was presumed to occur at the cortical level.

A SOFTWARE PACKAGE FOR UNSUPERVISED PATTERN RECOGNITION AND SYNOPTIC REPRESENTATION OF RESULTS: APPLICATION TO VOLCANIC TREMOR DATA OF MT ETNA

NASA Astrophysics Data System (ADS)

Langer, H. K.; Falsaperla, S. M.; Behncke, B.; Messina, A.; Spampinato, S.

2009-12-01

Artificial Intelligence (AI) has found broad applications in volcano observatories worldwide with the aim of reducing volcanic hazard. The need to process larger and larger quantity of data makes indeed AI techniques appealing for monitoring purposes. Tools based on Artificial Neural Networks and Support Vector Machine have proved to be particularly successful in the classification of seismic events and volcanic tremor changes heralding eruptive activity, such as paroxysmal explosions and lava fountaining at Stromboli and Mt Etna, Italy (e.g., Falsaperla et al., 1996; Langer et al., 2009). Moving on from the excellent results obtained from these applications, we present KKAnalysis, a MATLAB based software which combines several unsupervised pattern classification methods, exploiting routines of the SOM Toolbox 2 for MATLAB (http://www.cis.hut.fi/projects/somtoolbox). KKAnalysis is based on Self Organizing Maps (SOM) and clustering methods consisting of K-Means, Fuzzy C-Means, and a scheme based on a metrics accounting for correlation between components of the feature vector. We show examples of applications of this tool to volcanic tremor data recorded at Mt Etna between 2007 and 2009. This time span - during which Strombolian explosions, 7 episodes of lava fountaining and effusive activity occurred - is particularly interesting, as it encompassed different states of volcanic activity (i.e., non-eruptive, eruptive according to different styles) for the unsupervised classifier to identify, highlighting their development in time. Even subtle changes in the signal characteristics allow the unsupervised classifier to recognize features belonging to the different classes and stages of volcanic activity. A convenient color-code representation shows up the temporal development of the different classes of signal, making this method extremely helpful for monitoring purposes and surveillance. Though being developed for volcanic tremor classification, KKAnalysis is generally applicable to any type of physical or chemical pattern, provided that feature vectors are given in numerical form. References: Falsaperla, S., S. Graziani, G. Nunnari, and S. Spampinato (1996). Automatic classification of volcanic earthquakes by using multy-layered neural networks. Natural Hazard, 13, 205-228. Langer, H., S. Falsaperla, M. Masotti, R. Campanini, S. Spampinato, and A. Messina (2008). Synopsis of supervised and unsupervised pattern classification techniques applied to volcanic tremor data at Mt Etna, Italy. Geophys. J. Int., doi:10.1111/j.1365-246X.2009.04179.x.

Performance analysis of unsupervised optimal fuzzy clustering algorithm for MRI brain tumor segmentation.

PubMed

Blessy, S A Praylin Selva; Sulochana, C Helen

2015-01-01

Segmentation of brain tumor from Magnetic Resonance Imaging (MRI) becomes very complicated due to the structural complexities of human brain and the presence of intensity inhomogeneities. To propose a method that effectively segments brain tumor from MR images and to evaluate the performance of unsupervised optimal fuzzy clustering (UOFC) algorithm for segmentation of brain tumor from MR images. Segmentation is done by preprocessing the MR image to standardize intensity inhomogeneities followed by feature extraction, feature fusion and clustering. Different validation measures are used to evaluate the performance of the proposed method using different clustering algorithms. The proposed method using UOFC algorithm produces high sensitivity (96%) and low specificity (4%) compared to other clustering methods. Validation results clearly show that the proposed method with UOFC algorithm effectively segments brain tumor from MR images.

A consensus embedding approach for segmentation of high resolution in vivo prostate magnetic resonance imagery

NASA Astrophysics Data System (ADS)

Viswanath, Satish; Rosen, Mark; Madabhushi, Anant

2008-03-01

Current techniques for localization of prostatic adenocarcinoma (CaP) via blinded trans-rectal ultrasound biopsy are associated with a high false negative detection rate. While high resolution endorectal in vivo Magnetic Resonance (MR) prostate imaging has been shown to have improved contrast and resolution for CaP detection over ultrasound, similarity in intensity characteristics between benign and cancerous regions on MR images contribute to a high false positive detection rate. In this paper, we present a novel unsupervised segmentation method that employs manifold learning via consensus schemes for detection of cancerous regions from high resolution 1.5 Tesla (T) endorectal in vivo prostate MRI. A significant contribution of this paper is a method to combine multiple weak, lower-dimensional representations of high dimensional feature data in a way analogous to classifier ensemble schemes, and hence create a stable and accurate reduced dimensional representation. After correcting for MR image intensity artifacts, such as bias field inhomogeneity and intensity non-standardness, our algorithm extracts over 350 3D texture features at every spatial location in the MR scene at multiple scales and orientations. Non-linear dimensionality reduction schemes such as Locally Linear Embedding (LLE) and Graph Embedding (GE) are employed to create multiple low dimensional data representations of this high dimensional texture feature space. Our novel consensus embedding method is used to average object adjacencies from within the multiple low dimensional projections so that class relationships are preserved. Unsupervised consensus clustering is then used to partition the objects in this consensus embedding space into distinct classes. Quantitative evaluation on 18 1.5 T prostate MR data against corresponding histology obtained from the multi-site ACRIN trials show a sensitivity of 92.65% and a specificity of 82.06%, which suggests that our method is successfully able to detect suspicious regions in the prostate.

An Event-Driven Classifier for Spiking Neural Networks Fed with Synthetic or Dynamic Vision Sensor Data.

PubMed

Stromatias, Evangelos; Soto, Miguel; Serrano-Gotarredona, Teresa; Linares-Barranco, Bernabé

2017-01-01

This paper introduces a novel methodology for training an event-driven classifier within a Spiking Neural Network (SNN) System capable of yielding good classification results when using both synthetic input data and real data captured from Dynamic Vision Sensor (DVS) chips. The proposed supervised method uses the spiking activity provided by an arbitrary topology of prior SNN layers to build histograms and train the classifier in the frame domain using the stochastic gradient descent algorithm. In addition, this approach can cope with leaky integrate-and-fire neuron models within the SNN, a desirable feature for real-world SNN applications, where neural activation must fade away after some time in the absence of inputs. Consequently, this way of building histograms captures the dynamics of spikes immediately before the classifier. We tested our method on the MNIST data set using different synthetic encodings and real DVS sensory data sets such as N-MNIST, MNIST-DVS, and Poker-DVS using the same network topology and feature maps. We demonstrate the effectiveness of our approach by achieving the highest classification accuracy reported on the N-MNIST (97.77%) and Poker-DVS (100%) real DVS data sets to date with a spiking convolutional network. Moreover, by using the proposed method we were able to retrain the output layer of a previously reported spiking neural network and increase its performance by 2%, suggesting that the proposed classifier can be used as the output layer in works where features are extracted using unsupervised spike-based learning methods. In addition, we also analyze SNN performance figures such as total event activity and network latencies, which are relevant for eventual hardware implementations. In summary, the paper aggregates unsupervised-trained SNNs with a supervised-trained SNN classifier, combining and applying them to heterogeneous sets of benchmarks, both synthetic and from real DVS chips.

Deep Supervised, but Not Unsupervised, Models May Explain IT Cortical Representation

PubMed Central

Khaligh-Razavi, Seyed-Mahdi; Kriegeskorte, Nikolaus

2014-01-01

Inferior temporal (IT) cortex in human and nonhuman primates serves visual object recognition. Computational object-vision models, although continually improving, do not yet reach human performance. It is unclear to what extent the internal representations of computational models can explain the IT representation. Here we investigate a wide range of computational model representations (37 in total), testing their categorization performance and their ability to account for the IT representational geometry. The models include well-known neuroscientific object-recognition models (e.g. HMAX, VisNet) along with several models from computer vision (e.g. SIFT, GIST, self-similarity features, and a deep convolutional neural network). We compared the representational dissimilarity matrices (RDMs) of the model representations with the RDMs obtained from human IT (measured with fMRI) and monkey IT (measured with cell recording) for the same set of stimuli (not used in training the models). Better performing models were more similar to IT in that they showed greater clustering of representational patterns by category. In addition, better performing models also more strongly resembled IT in terms of their within-category representational dissimilarities. Representational geometries were significantly correlated between IT and many of the models. However, the categorical clustering observed in IT was largely unexplained by the unsupervised models. The deep convolutional network, which was trained by supervision with over a million category-labeled images, reached the highest categorization performance and also best explained IT, although it did not fully explain the IT data. Combining the features of this model with appropriate weights and adding linear combinations that maximize the margin between animate and inanimate objects and between faces and other objects yielded a representation that fully explained our IT data. Overall, our results suggest that explaining IT requires computational features trained through supervised learning to emphasize the behaviorally important categorical divisions prominently reflected in IT. PMID:25375136

Multispectral and Panchromatic used Enhancement Resolution and Study Effective Enhancement on Supervised and Unsupervised Classification Land – Cover

NASA Astrophysics Data System (ADS)

Salman, S. S.; Abbas, W. A.

2018-05-01

The goal of the study is to support analysis Enhancement of Resolution and study effect on classification methods on bands spectral information of specific and quantitative approaches. In this study introduce a method to enhancement resolution Landsat 8 of combining the bands spectral of 30 meters resolution with panchromatic band 8 of 15 meters resolution, because of importance multispectral imagery to extracting land - cover. Classification methods used in this study to classify several lands -covers recorded from OLI- 8 imagery. Two methods of Data mining can be classified as either supervised or unsupervised. In supervised methods, there is a particular predefined target, that means the algorithm learn which values of the target are associated with which values of the predictor sample. K-nearest neighbors and maximum likelihood algorithms examine in this work as supervised methods. In other hand, no sample identified as target in unsupervised methods, the algorithm of data extraction searches for structure and patterns between all the variables, represented by Fuzzy C-mean clustering method as one of the unsupervised methods, NDVI vegetation index used to compare the results of classification method, the percent of dense vegetation in maximum likelihood method give a best results.

Unsupervised categorization with individuals diagnosed as having moderate traumatic brain injury: Over-selective responding.

PubMed

Edwards, Darren J; Wood, Rodger

2016-01-01

This study explored over-selectivity (executive dysfunction) using a standard unsupervised categorization task. Over-selectivity has been demonstrated using supervised categorization procedures (where training is given); however, little has been done in the way of unsupervised categorization (without training). A standard unsupervised categorization task was used to assess levels of over-selectivity in a traumatic brain injury (TBI) population. Individuals with TBI were selected from the Tertiary Traumatic Brain Injury Clinic at Swansea University and were asked to categorize two-dimensional items (pictures on cards), into groups that they felt were most intuitive, and without any learning (feedback from experimenter). This was compared against categories made by a control group for the same task. The findings of this study demonstrate that individuals with TBI had deficits for both easy and difficult categorization sets, as indicated by a larger amount of one-dimensional sorting compared to control participants. Deficits were significantly greater for the easy condition. The implications of these findings are discussed in the context of over-selectivity, and the processes that underlie this deficit. Also, the implications for using this procedure as a screening measure for over-selectivity in TBI are discussed.

Segmentation of fluorescence microscopy cell images using unsupervised mining.

PubMed

Du, Xian; Dua, Sumeet

2010-05-28

The accurate measurement of cell and nuclei contours are critical for the sensitive and specific detection of changes in normal cells in several medical informatics disciplines. Within microscopy, this task is facilitated using fluorescence cell stains, and segmentation is often the first step in such approaches. Due to the complex nature of cell issues and problems inherent to microscopy, unsupervised mining approaches of clustering can be incorporated in the segmentation of cells. In this study, we have developed and evaluated the performance of multiple unsupervised data mining techniques in cell image segmentation. We adapt four distinctive, yet complementary, methods for unsupervised learning, including those based on k-means clustering, EM, Otsu's threshold, and GMAC. Validation measures are defined, and the performance of the techniques is evaluated both quantitatively and qualitatively using synthetic and recently published real data. Experimental results demonstrate that k-means, Otsu's threshold, and GMAC perform similarly, and have more precise segmentation results than EM. We report that EM has higher recall values and lower precision results from under-segmentation due to its Gaussian model assumption. We also demonstrate that these methods need spatial information to segment complex real cell images with a high degree of efficacy, as expected in many medical informatics applications.

An unsupervised machine learning method for delineating stratum corneum in reflectance confocal microscopy stacks of human skin in vivo

NASA Astrophysics Data System (ADS)

Bozkurt, Alican; Kose, Kivanc; Fox, Christi A.; Dy, Jennifer; Brooks, Dana H.; Rajadhyaksha, Milind

2016-02-01

Study of the stratum corneum (SC) in human skin is important for research in barrier structure and function, drug delivery, and water permeability of skin. The optical sectioning and high resolution of reflectance confocal microscopy (RCM) allows visual examination of SC non-invasively. Here, we present an unsupervised segmentation algorithm that can automatically delineate thickness of the SC in RCM images of human skin in-vivo. We mimic clinicians visual process by applying complex wavelet transform over non-overlapping local regions of size 16 x 16 μm called tiles, and analyze the textural changes in between consecutive tiles in axial (depth) direction. We use dual-tree complex wavelet transform to represent textural structures in each tile. This transform is almost shift-invariant, and directionally selective, which makes it highly efficient in texture representation. Using DT-CWT, we decompose each tile into 6 directional sub-bands with orientations in +/-15, 45, and 75 degrees and a low-pass band, which is the decimated version of the input. We apply 3 scales of decomposition by recursively transforming the low-pass bands and obtain 18 bands of different directionality at different scales. We then calculate mean and variance of each band resulting in a feature vector of 36 entries. Feature vectors obtained for each stack of tiles in axial direction are then clustered using spectral clustering in order to detect the textural changes in depth direction. Testing on a set of 15 RCM stacks produced a mean error of 5.45+/-1.32 μm, compared to the "ground truth" segmentation provided by a clinical expert reader.

Learning the 3-D structure of objects from 2-D views depends on shape, not format

PubMed Central

Tian, Moqian; Yamins, Daniel; Grill-Spector, Kalanit

2016-01-01

Humans can learn to recognize new objects just from observing example views. However, it is unknown what structural information enables this learning. To address this question, we manipulated the amount of structural information given to subjects during unsupervised learning by varying the format of the trained views. We then tested how format affected participants' ability to discriminate similar objects across views that were rotated 90° apart. We found that, after training, participants' performance increased and generalized to new views in the same format. Surprisingly, the improvement was similar across line drawings, shape from shading, and shape from shading + stereo even though the latter two formats provide richer depth information compared to line drawings. In contrast, participants' improvement was significantly lower when training used silhouettes, suggesting that silhouettes do not have enough information to generate a robust 3-D structure. To test whether the learned object representations were format-specific or format-invariant, we examined if learning novel objects from example views transfers across formats. We found that learning objects from example line drawings transferred to shape from shading and vice versa. These results have important implications for theories of object recognition because they suggest that (a) learning the 3-D structure of objects does not require rich structural cues during training as long as shape information of internal and external features is provided and (b) learning generates shape-based object representations independent of the training format. PMID:27153196

Chromatin accessibility prediction via convolutional long short-term memory networks with k-mer embedding.

PubMed

Min, Xu; Zeng, Wanwen; Chen, Ning; Chen, Ting; Jiang, Rui

2017-07-15

Experimental techniques for measuring chromatin accessibility are expensive and time consuming, appealing for the development of computational approaches to predict open chromatin regions from DNA sequences. Along this direction, existing methods fall into two classes: one based on handcrafted k -mer features and the other based on convolutional neural networks. Although both categories have shown good performance in specific applications thus far, there still lacks a comprehensive framework to integrate useful k -mer co-occurrence information with recent advances in deep learning. We fill this gap by addressing the problem of chromatin accessibility prediction with a convolutional Long Short-Term Memory (LSTM) network with k -mer embedding. We first split DNA sequences into k -mers and pre-train k -mer embedding vectors based on the co-occurrence matrix of k -mers by using an unsupervised representation learning approach. We then construct a supervised deep learning architecture comprised of an embedding layer, three convolutional layers and a Bidirectional LSTM (BLSTM) layer for feature learning and classification. We demonstrate that our method gains high-quality fixed-length features from variable-length sequences and consistently outperforms baseline methods. We show that k -mer embedding can effectively enhance model performance by exploring different embedding strategies. We also prove the efficacy of both the convolution and the BLSTM layers by comparing two variations of the network architecture. We confirm the robustness of our model to hyper-parameters by performing sensitivity analysis. We hope our method can eventually reinforce our understanding of employing deep learning in genomic studies and shed light on research regarding mechanisms of chromatin accessibility. The source code can be downloaded from https://github.com/minxueric/ismb2017_lstm . tingchen@tsinghua.edu.cn or ruijiang@tsinghua.edu.cn. Supplementary materials are available at Bioinformatics online. © The Author 2017. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com

Chromatin accessibility prediction via convolutional long short-term memory networks with k-mer embedding

PubMed Central

Min, Xu; Zeng, Wanwen; Chen, Ning; Chen, Ting; Jiang, Rui

2017-01-01

Abstract Motivation: Experimental techniques for measuring chromatin accessibility are expensive and time consuming, appealing for the development of computational approaches to predict open chromatin regions from DNA sequences. Along this direction, existing methods fall into two classes: one based on handcrafted k-mer features and the other based on convolutional neural networks. Although both categories have shown good performance in specific applications thus far, there still lacks a comprehensive framework to integrate useful k-mer co-occurrence information with recent advances in deep learning. Results: We fill this gap by addressing the problem of chromatin accessibility prediction with a convolutional Long Short-Term Memory (LSTM) network with k-mer embedding. We first split DNA sequences into k-mers and pre-train k-mer embedding vectors based on the co-occurrence matrix of k-mers by using an unsupervised representation learning approach. We then construct a supervised deep learning architecture comprised of an embedding layer, three convolutional layers and a Bidirectional LSTM (BLSTM) layer for feature learning and classification. We demonstrate that our method gains high-quality fixed-length features from variable-length sequences and consistently outperforms baseline methods. We show that k-mer embedding can effectively enhance model performance by exploring different embedding strategies. We also prove the efficacy of both the convolution and the BLSTM layers by comparing two variations of the network architecture. We confirm the robustness of our model to hyper-parameters by performing sensitivity analysis. We hope our method can eventually reinforce our understanding of employing deep learning in genomic studies and shed light on research regarding mechanisms of chromatin accessibility. Availability and implementation: The source code can be downloaded from https://github.com/minxueric/ismb2017_lstm. Contact: tingchen@tsinghua.edu.cn or ruijiang@tsinghua.edu.cn Supplementary information: Supplementary materials are available at Bioinformatics online. PMID:28881969

Correlation of HIV protease structure with Indinavir resistance: a data mining and neural networks approach

NASA Astrophysics Data System (ADS)

Draghici, Sorin; Cumberland, Lonnie T., Jr.; Kovari, Ladislau C.

2000-04-01

This paper presents some results of data mining HIV genotypic and structural data. Our aim is to try to relate structural features of HIV enzymes essential to its reproductive abilities to the drug resistance phenomenon. This paper concentrates on the HIV protease enzyme and Indinavir which is one of the FDA approved protease inhibitors. Our starting point was the current list of HIV mutations related to drug resistance. We used the fact that some molecular structures determined through high resolution X-ray crystallography were available for the protease-Indinavir complex. Starting with these structures and the known mutations, we modelled the mutant proteases and studied the pattern of atomic contacts between the protease and the drug. After suitable pre- processing, these patterns have been used as the input of our data mining process. We have used both supervised and unsupervised learning techniques with the aim of understanding the relationship between structural features at a molecular level and resistance to Indinavir. The supervised learning was aimed at predicting IC90 values for arbitrary mutants. The SOFM was aimed at identifying those structural features that are important for drug resistance and discovering a classifier based on such features. We have used validation and cross validation to test the generalization abilities of the learning paradigm we have designed. The straightforward supervised learning was able to learn very successfully but validation results are less than satisfactory. This is due to the insufficient number of patterns in the training set which in turn is due to the scarcity of the available data. The data mining using SOFM was very successful. We have managed to distinguish between resistant and non-resistant mutants using structural features. We have been able to divide all reported HIV mutants into several categories based on their 3- dimensional molecular structures and the pattern of contacts between the mutant protease and Indinavir. Our classifier shows reasonably good prediction performance being able to predict the drug resistance of previously unseen mutants with an accuracy of between 60% and 70%. We believe that this performance can be greatly improved once more data becomes available. The results presented here support the hypothesis that structural features of the molecular structure can be used in antiviral drug treatment selection and drug design.

Unsupervised domain adaptation for early detection of drought stress in hyperspectral images

NASA Astrophysics Data System (ADS)

Schmitter, P.; Steinrücken, J.; Römer, C.; Ballvora, A.; Léon, J.; Rascher, U.; Plümer, L.

2017-09-01

Hyperspectral images can be used to uncover physiological processes in plants if interpreted properly. Machine Learning methods such as Support Vector Machines (SVM) and Random Forests have been applied to estimate development of biomass and detect and predict plant diseases and drought stress. One basic requirement of machine learning implies, that training and testing is done in the same domain and the same distribution. Different genotypes, environmental conditions, illumination and sensors violate this requirement in most practical circumstances. Here, we present an approach, which enables the detection of physiological processes by transferring the prior knowledge within an existing model into a related target domain, where no label information is available. We propose a two-step transformation of the target features, which enables a direct application of an existing model. The transformation is evaluated by an objective function including additional prior knowledge about classification and physiological processes in plants. We have applied the approach to three sets of hyperspectral images, which were acquired with different plant species in different environments observed with different sensors. It is shown, that a classification model, derived on one of the sets, delivers satisfying classification results on the transformed features of the other data sets. Furthermore, in all cases early non-invasive detection of drought stress was possible.

Blind source computer device identification from recorded VoIP calls for forensic investigation.

PubMed

Jahanirad, Mehdi; Anuar, Nor Badrul; Wahab, Ainuddin Wahid Abdul

2017-03-01

The VoIP services provide fertile ground for criminal activity, thus identifying the transmitting computer devices from recorded VoIP call may help the forensic investigator to reveal useful information. It also proves the authenticity of the call recording submitted to the court as evidence. This paper extended the previous study on the use of recorded VoIP call for blind source computer device identification. Although initial results were promising but theoretical reasoning for this is yet to be found. The study suggested computing entropy of mel-frequency cepstrum coefficients (entropy-MFCC) from near-silent segments as an intrinsic feature set that captures the device response function due to the tolerances in the electronic components of individual computer devices. By applying the supervised learning techniques of naïve Bayesian, linear logistic regression, neural networks and support vector machines to the entropy-MFCC features, state-of-the-art identification accuracy of near 99.9% has been achieved on different sets of computer devices for both call recording and microphone recording scenarios. Furthermore, unsupervised learning techniques, including simple k-means, expectation-maximization and density-based spatial clustering of applications with noise (DBSCAN) provided promising results for call recording dataset by assigning the majority of instances to their correct clusters. Copyright © 2017 Elsevier Ireland Ltd. All rights reserved.

Stochastic Synapses Enable Efficient Brain-Inspired Learning Machines.

PubMed

Neftci, Emre O; Pedroni, Bruno U; Joshi, Siddharth; Al-Shedivat, Maruan; Cauwenberghs, Gert

2016-01-01

Recent studies have shown that synaptic unreliability is a robust and sufficient mechanism for inducing the stochasticity observed in cortex. Here, we introduce Synaptic Sampling Machines (S2Ms), a class of neural network models that uses synaptic stochasticity as a means to Monte Carlo sampling and unsupervised learning. Similar to the original formulation of Boltzmann machines, these models can be viewed as a stochastic counterpart of Hopfield networks, but where stochasticity is induced by a random mask over the connections. Synaptic stochasticity plays the dual role of an efficient mechanism for sampling, and a regularizer during learning akin to DropConnect. A local synaptic plasticity rule implementing an event-driven form of contrastive divergence enables the learning of generative models in an on-line fashion. S2Ms perform equally well using discrete-timed artificial units (as in Hopfield networks) or continuous-timed leaky integrate and fire neurons. The learned representations are remarkably sparse and robust to reductions in bit precision and synapse pruning: removal of more than 75% of the weakest connections followed by cursory re-learning causes a negligible performance loss on benchmark classification tasks. The spiking neuron-based S2Ms outperform existing spike-based unsupervised learners, while potentially offering substantial advantages in terms of power and complexity, and are thus promising models for on-line learning in brain-inspired hardware.

Stochastic Synapses Enable Efficient Brain-Inspired Learning Machines

PubMed Central

Neftci, Emre O.; Pedroni, Bruno U.; Joshi, Siddharth; Al-Shedivat, Maruan; Cauwenberghs, Gert

2016-01-01

Recent studies have shown that synaptic unreliability is a robust and sufficient mechanism for inducing the stochasticity observed in cortex. Here, we introduce Synaptic Sampling Machines (S2Ms), a class of neural network models that uses synaptic stochasticity as a means to Monte Carlo sampling and unsupervised learning. Similar to the original formulation of Boltzmann machines, these models can be viewed as a stochastic counterpart of Hopfield networks, but where stochasticity is induced by a random mask over the connections. Synaptic stochasticity plays the dual role of an efficient mechanism for sampling, and a regularizer during learning akin to DropConnect. A local synaptic plasticity rule implementing an event-driven form of contrastive divergence enables the learning of generative models in an on-line fashion. S2Ms perform equally well using discrete-timed artificial units (as in Hopfield networks) or continuous-timed leaky integrate and fire neurons. The learned representations are remarkably sparse and robust to reductions in bit precision and synapse pruning: removal of more than 75% of the weakest connections followed by cursory re-learning causes a negligible performance loss on benchmark classification tasks. The spiking neuron-based S2Ms outperform existing spike-based unsupervised learners, while potentially offering substantial advantages in terms of power and complexity, and are thus promising models for on-line learning in brain-inspired hardware. PMID:27445650

An Unsupervised kNN Method to Systematically Detect Changes in Protein Localization in High-Throughput Microscopy Images.

PubMed

Lu, Alex Xijie; Moses, Alan M

2016-01-01

Despite the importance of characterizing genes that exhibit subcellular localization changes between conditions in proteome-wide imaging experiments, many recent studies still rely upon manual evaluation to assess the results of high-throughput imaging experiments. We describe and demonstrate an unsupervised k-nearest neighbours method for the detection of localization changes. Compared to previous classification-based supervised change detection methods, our method is much simpler and faster, and operates directly on the feature space to overcome limitations in needing to manually curate training sets that may not generalize well between screens. In addition, the output of our method is flexible in its utility, generating both a quantitatively ranked list of localization changes that permit user-defined cut-offs, and a vector for each gene describing feature-wise direction and magnitude of localization changes. We demonstrate that our method is effective at the detection of localization changes using the Δrpd3 perturbation in Saccharomyces cerevisiae, where we capture 71.4% of previously known changes within the top 10% of ranked genes, and find at least four new localization changes within the top 1% of ranked genes. The results of our analysis indicate that simple unsupervised methods may be able to identify localization changes in images without laborious manual image labelling steps.

Selective Convolutional Descriptor Aggregation for Fine-Grained Image Retrieval.

PubMed

Wei, Xiu-Shen; Luo, Jian-Hao; Wu, Jianxin; Zhou, Zhi-Hua

2017-06-01

Deep convolutional neural network models pre-trained for the ImageNet classification task have been successfully adopted to tasks in other domains, such as texture description and object proposal generation, but these tasks require annotations for images in the new domain. In this paper, we focus on a novel and challenging task in the pure unsupervised setting: fine-grained image retrieval. Even with image labels, fine-grained images are difficult to classify, letting alone the unsupervised retrieval task. We propose the selective convolutional descriptor aggregation (SCDA) method. The SCDA first localizes the main object in fine-grained images, a step that discards the noisy background and keeps useful deep descriptors. The selected descriptors are then aggregated and the dimensionality is reduced into a short feature vector using the best practices we found. The SCDA is unsupervised, using no image label or bounding box annotation. Experiments on six fine-grained data sets confirm the effectiveness of the SCDA for fine-grained image retrieval. Besides, visualization of the SCDA features shows that they correspond to visual attributes (even subtle ones), which might explain SCDA's high-mean average precision in fine-grained retrieval. Moreover, on general image retrieval data sets, the SCDA achieves comparable retrieval results with the state-of-the-art general image retrieval approaches.

Characterizing Interference in Radio Astronomy Observations through Active and Unsupervised Learning

NASA Technical Reports Server (NTRS)

Doran, G.

2013-01-01

In the process of observing signals from astronomical sources, radio astronomers must mitigate the effects of manmade radio sources such as cell phones, satellites, aircraft, and observatory equipment. Radio frequency interference (RFI) often occurs as short bursts (< 1 ms) across a broad range of frequencies, and can be confused with signals from sources of interest such as pulsars. With ever-increasing volumes of data being produced by observatories, automated strategies are required to detect, classify, and characterize these short "transient" RFI events. We investigate an active learning approach in which an astronomer labels events that are most confusing to a classifier, minimizing the human effort required for classification. We also explore the use of unsupervised clustering techniques, which automatically group events into classes without user input. We apply these techniques to data from the Parkes Multibeam Pulsar Survey to characterize several million detected RFI events from over a thousand hours of observation.

Unsupervised segmentation with dynamical units.

PubMed

Rao, A Ravishankar; Cecchi, Guillermo A; Peck, Charles C; Kozloski, James R

2008-01-01

In this paper, we present a novel network to separate mixtures of inputs that have been previously learned. A significant capability of the network is that it segments the components of each input object that most contribute to its classification. The network consists of amplitude-phase units that can synchronize their dynamics, so that separation is determined by the amplitude of units in an output layer, and segmentation by phase similarity between input and output layer units. Learning is unsupervised and based on a Hebbian update, and the architecture is very simple. Moreover, efficient segmentation can be achieved even when there is considerable superposition of the inputs. The network dynamics are derived from an objective function that rewards sparse coding in the generalized amplitude-phase variables. We argue that this objective function can provide a possible formal interpretation of the binding problem and that the implementation of the network architecture and dynamics is biologically plausible.

Active Learning with Rationales for Identifying Operationally Significant Anomalies in Aviation

NASA Technical Reports Server (NTRS)

Sharma, Manali; Das, Kamalika; Bilgic, Mustafa; Matthews, Bryan; Nielsen, David Lynn; Oza, Nikunj C.

2016-01-01

A major focus of the commercial aviation community is discovery of unknown safety events in flight operations data. Data-driven unsupervised anomaly detection methods are better at capturing unknown safety events compared to rule-based methods which only look for known violations. However, not all statistical anomalies that are discovered by these unsupervised anomaly detection methods are operationally significant (e.g., represent a safety concern). Subject Matter Experts (SMEs) have to spend significant time reviewing these statistical anomalies individually to identify a few operationally significant ones. In this paper we propose an active learning algorithm that incorporates SME feedback in the form of rationales to build a classifier that can distinguish between uninteresting and operationally significant anomalies. Experimental evaluation on real aviation data shows that our approach improves detection of operationally significant events by as much as 75% compared to the state-of-the-art. The learnt classifier also generalizes well to additional validation data sets.

Stratification Learning: Detecting Mixed Density and Dimensionality in High Dimensional Point Clouds (PREPRINT)

DTIC Science & Technology

2006-09-01

Medioni, [11], estimates the local dimension using tensor voting . These recent works have clearly shown the necessity to go beyond manifold learning, into...2005. [11] P. Mordohai and G. Medioni. Unsupervised dimensionality estimation and manifold learning in high-dimensional spaces by tensor voting . In...walking, jumping, and arms waving. The whole run took 361 seconds in Matlab , while the classification time (PMM) can be neglected compared to the kNN

How to Select a Good Training-data Subset for Transcription: Submodular Active Selection for Sequences

DTIC Science & Technology

2009-01-01

selection and uncertainty sampling signif- icantly. Index Terms: Transcription, labeling, submodularity, submod- ular selection, active learning , sequence...name of batch active learning , where a subset of data that is most informative and represen- tative of the whole is selected for labeling. Often...representative subset. Note that our Fisher ker- nel is over an unsupervised generative model, which enables us to bootstrap our active learning approach

Unsupervised segmentation of H and E breast images

NASA Astrophysics Data System (ADS)

Hope, Tyna A.; Yaffe, Martin J.

2017-03-01

Heterogeneity of ductal carcinoma in situ (DCIS) continues to be an important topic. Combining biomarker and hematoxylin and eosin (HE) morphology information may provide more insights than either alone. We are working towards a computer-based identification and description system for DCIS. As part of the system we are developing a region of interest finder for further processing, such as identifying DCIS and other HE based measures. The segmentation algorithm is designed to be tolerant of variability in staining and require no user interaction. To achieve stain variation tolerance we use unsupervised learning and iteratively interrogate the image for information. Using simple rules (e.g., "hematoxylin stains nuclei") and iteratively assessing the resultant objects (small hematoxylin stained objects are lymphocytes), the system builds up a knowledge base so that it is not dependent upon manual annotations. The system starts with image resolution-based assumptions but these are replaced by knowledge gained. The algorithm pipeline is designed to find the simplest items first (segment stains), then interesting subclasses and objects (stroma, lymphocytes), and builds information until it is possible to segment blobs that are normal, DCIS, and the range of benign glands. Once the blobs are found, features can be obtained and DCIS detected. In this work we present the early segmentation results with stains where hematoxylin ranges from blue dominant to red dominant in RGB space.

Spectral Transfer Learning Using Information Geometry for a User-Independent Brain-Computer Interface

PubMed Central

Waytowich, Nicholas R.; Lawhern, Vernon J.; Bohannon, Addison W.; Ball, Kenneth R.; Lance, Brent J.

2016-01-01

Recent advances in signal processing and machine learning techniques have enabled the application of Brain-Computer Interface (BCI) technologies to fields such as medicine, industry, and recreation; however, BCIs still suffer from the requirement of frequent calibration sessions due to the intra- and inter-individual variability of brain-signals, which makes calibration suppression through transfer learning an area of increasing interest for the development of practical BCI systems. In this paper, we present an unsupervised transfer method (spectral transfer using information geometry, STIG), which ranks and combines unlabeled predictions from an ensemble of information geometry classifiers built on data from individual training subjects. The STIG method is validated in both off-line and real-time feedback analysis during a rapid serial visual presentation task (RSVP). For detection of single-trial, event-related potentials (ERPs), the proposed method can significantly outperform existing calibration-free techniques as well as outperform traditional within-subject calibration techniques when limited data is available. This method demonstrates that unsupervised transfer learning for single-trial detection in ERP-based BCIs can be achieved without the requirement of costly training data, representing a step-forward in the overall goal of achieving a practical user-independent BCI system. PMID:27713685

Spectral Transfer Learning Using Information Geometry for a User-Independent Brain-Computer Interface.

PubMed

Waytowich, Nicholas R; Lawhern, Vernon J; Bohannon, Addison W; Ball, Kenneth R; Lance, Brent J

2016-01-01

Recent advances in signal processing and machine learning techniques have enabled the application of Brain-Computer Interface (BCI) technologies to fields such as medicine, industry, and recreation; however, BCIs still suffer from the requirement of frequent calibration sessions due to the intra- and inter-individual variability of brain-signals, which makes calibration suppression through transfer learning an area of increasing interest for the development of practical BCI systems. In this paper, we present an unsupervised transfer method (spectral transfer using information geometry, STIG), which ranks and combines unlabeled predictions from an ensemble of information geometry classifiers built on data from individual training subjects. The STIG method is validated in both off-line and real-time feedback analysis during a rapid serial visual presentation task (RSVP). For detection of single-trial, event-related potentials (ERPs), the proposed method can significantly outperform existing calibration-free techniques as well as outperform traditional within-subject calibration techniques when limited data is available. This method demonstrates that unsupervised transfer learning for single-trial detection in ERP-based BCIs can be achieved without the requirement of costly training data, representing a step-forward in the overall goal of achieving a practical user-independent BCI system.

Automatic microseismic event picking via unsupervised machine learning

NASA Astrophysics Data System (ADS)

Chen, Yangkang

2018-01-01

Effective and efficient arrival picking plays an important role in microseismic and earthquake data processing and imaging. Widely used short-term-average long-term-average ratio (STA/LTA) based arrival picking algorithms suffer from the sensitivity to moderate-to-strong random ambient noise. To make the state-of-the-art arrival picking approaches effective, microseismic data need to be first pre-processed, for example, removing sufficient amount of noise, and second analysed by arrival pickers. To conquer the noise issue in arrival picking for weak microseismic or earthquake event, I leverage the machine learning techniques to help recognizing seismic waveforms in microseismic or earthquake data. Because of the dependency of supervised machine learning algorithm on large volume of well-designed training data, I utilize an unsupervised machine learning algorithm to help cluster the time samples into two groups, that is, waveform points and non-waveform points. The fuzzy clustering algorithm has been demonstrated to be effective for such purpose. A group of synthetic, real microseismic and earthquake data sets with different levels of complexity show that the proposed method is much more robust than the state-of-the-art STA/LTA method in picking microseismic events, even in the case of moderately strong background noise.

Memristive neural network for on-line learning and tracking with brain-inspired spike timing dependent plasticity.

PubMed

Pedretti, G; Milo, V; Ambrogio, S; Carboni, R; Bianchi, S; Calderoni, A; Ramaswamy, N; Spinelli, A S; Ielmini, D

2017-07-13

Brain-inspired computation can revolutionize information technology by introducing machines capable of recognizing patterns (images, speech, video) and interacting with the external world in a cognitive, humanlike way. Achieving this goal requires first to gain a detailed understanding of the brain operation, and second to identify a scalable microelectronic technology capable of reproducing some of the inherent functions of the human brain, such as the high synaptic connectivity (~10 4 ) and the peculiar time-dependent synaptic plasticity. Here we demonstrate unsupervised learning and tracking in a spiking neural network with memristive synapses, where synaptic weights are updated via brain-inspired spike timing dependent plasticity (STDP). The synaptic conductance is updated by the local time-dependent superposition of pre- and post-synaptic spikes within a hybrid one-transistor/one-resistor (1T1R) memristive synapse. Only 2 synaptic states, namely the low resistance state (LRS) and the high resistance state (HRS), are sufficient to learn and recognize patterns. Unsupervised learning of a static pattern and tracking of a dynamic pattern of up to 4 × 4 pixels are demonstrated, paving the way for intelligent hardware technology with up-scaled memristive neural networks.

Automated egg grading system using computer vision: Investigation on weight measure versus shape parameters

NASA Astrophysics Data System (ADS)

Nasir, Ahmad Fakhri Ab; Suhaila Sabarudin, Siti; Majeed, Anwar P. P. Abdul; Ghani, Ahmad Shahrizan Abdul

2018-04-01

Chicken egg is a source of food of high demand by humans. Human operators cannot work perfectly and continuously when conducting egg grading. Instead of an egg grading system using weight measure, an automatic system for egg grading using computer vision (using egg shape parameter) can be used to improve the productivity of egg grading. However, early hypothesis has indicated that more number of egg classes will change when using egg shape parameter compared with using weight measure. This paper presents the comparison of egg classification by the two above-mentioned methods. Firstly, 120 images of chicken eggs of various grades (A–D) produced in Malaysia are captured. Then, the egg images are processed using image pre-processing techniques, such as image cropping, smoothing and segmentation. Thereafter, eight egg shape features, including area, major axis length, minor axis length, volume, diameter and perimeter, are extracted. Lastly, feature selection (information gain ratio) and feature extraction (principal component analysis) are performed using k-nearest neighbour classifier in the classification process. Two methods, namely, supervised learning (using weight measure as graded by egg supplier) and unsupervised learning (using egg shape parameters as graded by ourselves), are conducted to execute the experiment. Clustering results reveal many changes in egg classes after performing shape-based grading. On average, the best recognition results using shape-based grading label is 94.16% while using weight-based label is 44.17%. As conclusion, automated egg grading system using computer vision is better by implementing shape-based features since it uses image meanwhile the weight parameter is more suitable by using weight grading system.

Hierarchical Representation Learning for Kinship Verification.

PubMed

Kohli, Naman; Vatsa, Mayank; Singh, Richa; Noore, Afzel; Majumdar, Angshul

2017-01-01

Kinship verification has a number of applications such as organizing large collections of images and recognizing resemblances among humans. In this paper, first, a human study is conducted to understand the capabilities of human mind and to identify the discriminatory areas of a face that facilitate kinship-cues. The visual stimuli presented to the participants determine their ability to recognize kin relationship using the whole face as well as specific facial regions. The effect of participant gender and age and kin-relation pair of the stimulus is analyzed using quantitative measures such as accuracy, discriminability index d' , and perceptual information entropy. Utilizing the information obtained from the human study, a hierarchical kinship verification via representation learning (KVRL) framework is utilized to learn the representation of different face regions in an unsupervised manner. We propose a novel approach for feature representation termed as filtered contractive deep belief networks (fcDBN). The proposed feature representation encodes relational information present in images using filters and contractive regularization penalty. A compact representation of facial images of kin is extracted as an output from the learned model and a multi-layer neural network is utilized to verify the kin accurately. A new WVU kinship database is created, which consists of multiple images per subject to facilitate kinship verification. The results show that the proposed deep learning framework (KVRL-fcDBN) yields the state-of-the-art kinship verification accuracy on the WVU kinship database and on four existing benchmark data sets. Furthermore, kinship information is used as a soft biometric modality to boost the performance of face verification via product of likelihood ratio and support vector machine based approaches. Using the proposed KVRL-fcDBN framework, an improvement of over 20% is observed in the performance of face verification.

Unsupervised learning toward brain imaging data analysis: cigarette craving and resistance related neuronal activations from functional magnetic resonance imaging data analysis

NASA Astrophysics Data System (ADS)

Kim, Dong-Youl; Lee, Jong-Hwan

2014-05-01

A data-driven unsupervised learning such as an independent component analysis was gainfully applied to bloodoxygenation- level-dependent (BOLD) functional magnetic resonance imaging (fMRI) data compared to a model-based general linear model (GLM). This is due to an ability of this unsupervised learning method to extract a meaningful neuronal activity from BOLD signal that is a mixture of confounding non-neuronal artifacts such as head motions and physiological artifacts as well as neuronal signals. In this study, we support this claim by identifying neuronal underpinnings of cigarette craving and cigarette resistance. The fMRI data were acquired from heavy cigarette smokers (n = 14) while they alternatively watched images with and without cigarette smoking. During acquisition of two fMRI runs, they were asked to crave when they watched cigarette smoking images or to resist the urge to smoke. Data driven approaches of group independent component analysis (GICA) method based on temporal concatenation (TC) and TCGICA with an extension of iterative dual-regression (TC-GICA-iDR) were applied to the data. From the results, cigarette craving and cigarette resistance related neuronal activations were identified in the visual area and superior frontal areas, respectively with a greater statistical significance from the TC-GICA-iDR method than the TC-GICA method. On the other hand, the neuronal activity levels in many of these regions were not statistically different from the GLM method between the cigarette craving and cigarette resistance due to potentially aberrant BOLD signals.

Vibration control of building structures using self-organizing and self-learning neural networks

NASA Astrophysics Data System (ADS)

Madan, Alok

2005-11-01

Past research in artificial intelligence establishes that artificial neural networks (ANN) are effective and efficient computational processors for performing a variety of tasks including pattern recognition, classification, associative recall, combinatorial problem solving, adaptive control, multi-sensor data fusion, noise filtering and data compression, modelling and forecasting. The paper presents a potentially feasible approach for training ANN in active control of earthquake-induced vibrations in building structures without the aid of teacher signals (i.e. target control forces). A counter-propagation neural network is trained to output the control forces that are required to reduce the structural vibrations in the absence of any feedback on the correctness of the output control forces (i.e. without any information on the errors in output activations of the network). The present study shows that, in principle, the counter-propagation network (CPN) can learn from the control environment to compute the required control forces without the supervision of a teacher (unsupervised learning). Simulated case studies are presented to demonstrate the feasibility of implementing the unsupervised learning approach in ANN for effective vibration control of structures under the influence of earthquake ground motions. The proposed learning methodology obviates the need for developing a mathematical model of structural dynamics or training a separate neural network to emulate the structural response for implementation in practice.

Self-Organizing Feature Maps Identify Proteins Critical to Learning in a Mouse Model of Down Syndrome.

PubMed

Higuera, Clara; Gardiner, Katheleen J; Cios, Krzysztof J

2015-01-01

Down syndrome (DS) is a chromosomal abnormality (trisomy of human chromosome 21) associated with intellectual disability and affecting approximately one in 1000 live births worldwide. The overexpression of genes encoded by the extra copy of a normal chromosome in DS is believed to be sufficient to perturb normal pathways and normal responses to stimulation, causing learning and memory deficits. In this work, we have designed a strategy based on the unsupervised clustering method, Self Organizing Maps (SOM), to identify biologically important differences in protein levels in mice exposed to context fear conditioning (CFC). We analyzed expression levels of 77 proteins obtained from normal genotype control mice and from their trisomic littermates (Ts65Dn) both with and without treatment with the drug memantine. Control mice learn successfully while the trisomic mice fail, unless they are first treated with the drug, which rescues their learning ability. The SOM approach identified reduced subsets of proteins predicted to make the most critical contributions to normal learning, to failed learning and rescued learning, and provides a visual representation of the data that allows the user to extract patterns that may underlie novel biological responses to the different kinds of learning and the response to memantine. Results suggest that the application of SOM to new experimental data sets of complex protein profiles can be used to identify common critical protein responses, which in turn may aid in identifying potentially more effective drug targets.

Self-Organizing Feature Maps Identify Proteins Critical to Learning in a Mouse Model of Down Syndrome

PubMed Central

Higuera, Clara; Gardiner, Katheleen J.; Cios, Krzysztof J.

2015-01-01

Down syndrome (DS) is a chromosomal abnormality (trisomy of human chromosome 21) associated with intellectual disability and affecting approximately one in 1000 live births worldwide. The overexpression of genes encoded by the extra copy of a normal chromosome in DS is believed to be sufficient to perturb normal pathways and normal responses to stimulation, causing learning and memory deficits. In this work, we have designed a strategy based on the unsupervised clustering method, Self Organizing Maps (SOM), to identify biologically important differences in protein levels in mice exposed to context fear conditioning (CFC). We analyzed expression levels of 77 proteins obtained from normal genotype control mice and from their trisomic littermates (Ts65Dn) both with and without treatment with the drug memantine. Control mice learn successfully while the trisomic mice fail, unless they are first treated with the drug, which rescues their learning ability. The SOM approach identified reduced subsets of proteins predicted to make the most critical contributions to normal learning, to failed learning and rescued learning, and provides a visual representation of the data that allows the user to extract patterns that may underlie novel biological responses to the different kinds of learning and the response to memantine. Results suggest that the application of SOM to new experimental data sets of complex protein profiles can be used to identify common critical protein responses, which in turn may aid in identifying potentially more effective drug targets. PMID:26111164

Unsupervised semantic indoor scene classification for robot vision based on context of features using Gist and HSV-SIFT

NASA Astrophysics Data System (ADS)

Madokoro, H.; Yamanashi, A.; Sato, K.

2013-08-01

This paper presents an unsupervised scene classification method for actualizing semantic recognition of indoor scenes. Background and foreground features are respectively extracted using Gist and color scale-invariant feature transform (SIFT) as feature representations based on context. We used hue, saturation, and value SIFT (HSV-SIFT) because of its simple algorithm with low calculation costs. Our method creates bags of features for voting visual words created from both feature descriptors to a two-dimensional histogram. Moreover, our method generates labels as candidates of categories for time-series images while maintaining stability and plasticity together. Automatic labeling of category maps can be realized using labels created using adaptive resonance theory (ART) as teaching signals for counter propagation networks (CPNs). We evaluated our method for semantic scene classification using KTH's image database for robot localization (KTH-IDOL), which is popularly used for robot localization and navigation. The mean classification accuracies of Gist, gray SIFT, one class support vector machines (OC-SVM), position-invariant robust features (PIRF), and our method are, respectively, 39.7, 58.0, 56.0, 63.6, and 79.4%. The result of our method is 15.8% higher than that of PIRF. Moreover, we applied our method for fine classification using our original mobile robot. We obtained mean classification accuracy of 83.2% for six zones.

A parallelized binary search tree

USDA-ARS?s Scientific Manuscript database

PTTRNFNDR is an unsupervised statistical learning algorithm that detects patterns in DNA sequences, protein sequences, or any natural language texts that can be decomposed into letters of a finite alphabet. PTTRNFNDR performs complex mathematical computations and its processing time increases when i...

Selective Transfer Machine for Personalized Facial Action Unit Detection

PubMed Central

Chu, Wen-Sheng; De la Torre, Fernando; Cohn, Jeffery F.

2014-01-01

Automatic facial action unit (AFA) detection from video is a long-standing problem in facial expression analysis. Most approaches emphasize choices of features and classifiers. They neglect individual differences in target persons. People vary markedly in facial morphology (e.g., heavy versus delicate brows, smooth versus deeply etched wrinkles) and behavior. Individual differences can dramatically influence how well generic classifiers generalize to previously unseen persons. While a possible solution would be to train person-specific classifiers, that often is neither feasible nor theoretically compelling. The alternative that we propose is to personalize a generic classifier in an unsupervised manner (no additional labels for the test subjects are required). We introduce a transductive learning method, which we refer to Selective Transfer Machine (STM), to personalize a generic classifier by attenuating person-specific biases. STM achieves this effect by simultaneously learning a classifier and re-weighting the training samples that are most relevant to the test subject. To evaluate the effectiveness of STM, we compared STM to generic classifiers and to cross-domain learning methods in three major databases: CK+ [20], GEMEP-FERA [32] and RU-FACS [2]. STM outperformed generic classifiers in all. PMID:25242877

Unsupervised Metric Fusion Over Multiview Data by Graph Random Walk-Based Cross-View Diffusion.

PubMed

Wang, Yang; Zhang, Wenjie; Wu, Lin; Lin, Xuemin; Zhao, Xiang

2017-01-01

Learning an ideal metric is crucial to many tasks in computer vision. Diverse feature representations may combat this problem from different aspects; as visual data objects described by multiple features can be decomposed into multiple views, thus often provide complementary information. In this paper, we propose a cross-view fusion algorithm that leads to a similarity metric for multiview data by systematically fusing multiple similarity measures. Unlike existing paradigms, we focus on learning distance measure by exploiting a graph structure of data samples, where an input similarity matrix can be improved through a propagation of graph random walk. In particular, we construct multiple graphs with each one corresponding to an individual view, and a cross-view fusion approach based on graph random walk is presented to derive an optimal distance measure by fusing multiple metrics. Our method is scalable to a large amount of data by enforcing sparsity through an anchor graph representation. To adaptively control the effects of different views, we dynamically learn view-specific coefficients, which are leveraged into graph random walk to balance multiviews. However, such a strategy may lead to an over-smooth similarity metric where affinities between dissimilar samples may be enlarged by excessively conducting cross-view fusion. Thus, we figure out a heuristic approach to controlling the iteration number in the fusion process in order to avoid over smoothness. Extensive experiments conducted on real-world data sets validate the effectiveness and efficiency of our approach.

AHIMSA - Ad hoc histogram information measure sensing algorithm for feature selection in the context of histogram inspired clustering techniques

NASA Technical Reports Server (NTRS)

Dasarathy, B. V.

1976-01-01

An algorithm is proposed for dimensionality reduction in the context of clustering techniques based on histogram analysis. The approach is based on an evaluation of the hills and valleys in the unidimensional histograms along the different features and provides an economical means of assessing the significance of the features in a nonparametric unsupervised data environment. The method has relevance to remote sensing applications.

Semisupervised Clustering by Iterative Partition and Regression with Neuroscience Applications

PubMed Central

Qian, Guoqi; Wu, Yuehua; Ferrari, Davide; Qiao, Puxue; Hollande, Frédéric

2016-01-01

Regression clustering is a mixture of unsupervised and supervised statistical learning and data mining method which is found in a wide range of applications including artificial intelligence and neuroscience. It performs unsupervised learning when it clusters the data according to their respective unobserved regression hyperplanes. The method also performs supervised learning when it fits regression hyperplanes to the corresponding data clusters. Applying regression clustering in practice requires means of determining the underlying number of clusters in the data, finding the cluster label of each data point, and estimating the regression coefficients of the model. In this paper, we review the estimation and selection issues in regression clustering with regard to the least squares and robust statistical methods. We also provide a model selection based technique to determine the number of regression clusters underlying the data. We further develop a computing procedure for regression clustering estimation and selection. Finally, simulation studies are presented for assessing the procedure, together with analyzing a real data set on RGB cell marking in neuroscience to illustrate and interpret the method. PMID:27212939

Prototyping and Simulation of Robot Group Intelligence using Kohonen Networks.

PubMed

Wang, Zhijun; Mirdamadi, Reza; Wang, Qing

2016-01-01

Intelligent agents such as robots can form ad hoc networks and replace human being in many dangerous scenarios such as a complicated disaster relief site. This project prototypes and builds a computer simulator to simulate robot kinetics, unsupervised learning using Kohonen networks, as well as group intelligence when an ad hoc network is formed. Each robot is modeled using an object with a simple set of attributes and methods that define its internal states and possible actions it may take under certain circumstances. As the result, simple, reliable, and affordable robots can be deployed to form the network. The simulator simulates a group of robots as an unsupervised learning unit and tests the learning results under scenarios with different complexities. The simulation results show that a group of robots could demonstrate highly collaborative behavior on a complex terrain. This study could potentially provide a software simulation platform for testing individual and group capability of robots before the design process and manufacturing of robots. Therefore, results of the project have the potential to reduce the cost and improve the efficiency of robot design and building.

Topic Model for Graph Mining.

PubMed

Xuan, Junyu; Lu, Jie; Zhang, Guangquan; Luo, Xiangfeng

2015-12-01

Graph mining has been a popular research area because of its numerous application scenarios. Many unstructured and structured data can be represented as graphs, such as, documents, chemical molecular structures, and images. However, an issue in relation to current research on graphs is that they cannot adequately discover the topics hidden in graph-structured data which can be beneficial for both the unsupervised learning and supervised learning of the graphs. Although topic models have proved to be very successful in discovering latent topics, the standard topic models cannot be directly applied to graph-structured data due to the "bag-of-word" assumption. In this paper, an innovative graph topic model (GTM) is proposed to address this issue, which uses Bernoulli distributions to model the edges between nodes in a graph. It can, therefore, make the edges in a graph contribute to latent topic discovery and further improve the accuracy of the supervised and unsupervised learning of graphs. The experimental results on two different types of graph datasets show that the proposed GTM outperforms the latent Dirichlet allocation on classification by using the unveiled topics of these two models to represent graphs.

Prototyping and Simulation of Robot Group Intelligence using Kohonen Networks

PubMed Central

Wang, Zhijun; Mirdamadi, Reza; Wang, Qing

2016-01-01

Intelligent agents such as robots can form ad hoc networks and replace human being in many dangerous scenarios such as a complicated disaster relief site. This project prototypes and builds a computer simulator to simulate robot kinetics, unsupervised learning using Kohonen networks, as well as group intelligence when an ad hoc network is formed. Each robot is modeled using an object with a simple set of attributes and methods that define its internal states and possible actions it may take under certain circumstances. As the result, simple, reliable, and affordable robots can be deployed to form the network. The simulator simulates a group of robots as an unsupervised learning unit and tests the learning results under scenarios with different complexities. The simulation results show that a group of robots could demonstrate highly collaborative behavior on a complex terrain. This study could potentially provide a software simulation platform for testing individual and group capability of robots before the design process and manufacturing of robots. Therefore, results of the project have the potential to reduce the cost and improve the efficiency of robot design and building. PMID:28540284

Integrative Data Analysis of Multi-Platform Cancer Data with a Multimodal Deep Learning Approach.

PubMed

Liang, Muxuan; Li, Zhizhong; Chen, Ting; Zeng, Jianyang

2015-01-01

Identification of cancer subtypes plays an important role in revealing useful insights into disease pathogenesis and advancing personalized therapy. The recent development of high-throughput sequencing technologies has enabled the rapid collection of multi-platform genomic data (e.g., gene expression, miRNA expression, and DNA methylation) for the same set of tumor samples. Although numerous integrative clustering approaches have been developed to analyze cancer data, few of them are particularly designed to exploit both deep intrinsic statistical properties of each input modality and complex cross-modality correlations among multi-platform input data. In this paper, we propose a new machine learning model, called multimodal deep belief network (DBN), to cluster cancer patients from multi-platform observation data. In our integrative clustering framework, relationships among inherent features of each single modality are first encoded into multiple layers of hidden variables, and then a joint latent model is employed to fuse common features derived from multiple input modalities. A practical learning algorithm, called contrastive divergence (CD), is applied to infer the parameters of our multimodal DBN model in an unsupervised manner. Tests on two available cancer datasets show that our integrative data analysis approach can effectively extract a unified representation of latent features to capture both intra- and cross-modality correlations, and identify meaningful disease subtypes from multi-platform cancer data. In addition, our approach can identify key genes and miRNAs that may play distinct roles in the pathogenesis of different cancer subtypes. Among those key miRNAs, we found that the expression level of miR-29a is highly correlated with survival time in ovarian cancer patients. These results indicate that our multimodal DBN based data analysis approach may have practical applications in cancer pathogenesis studies and provide useful guidelines for personalized cancer therapy.

Chemometric analysis of correlations between electronic absorption characteristics and structural and/or physicochemical parameters for ampholytic substances of biological and pharmaceutical relevance.

PubMed

Judycka-Proma, U; Bober, L; Gajewicz, A; Puzyn, T; Błażejowski, J

2015-03-05

Forty ampholytic compounds of biological and pharmaceutical relevance were subjected to chemometric analysis based on unsupervised and supervised learning algorithms. This enabled relations to be found between empirical spectral characteristics derived from electronic absorption data and structural and physicochemical parameters predicted by quantum chemistry methods or phenomenological relationships based on additivity rules. It was found that the energies of long wavelength absorption bands are correlated through multiparametric linear relationships with parameters reflecting the bulkiness features of the absorbing molecules as well as their nucleophilicity and electrophilicity. These dependences enable the quantitative analysis of spectral features of the compounds, as well as a comparison of their similarities and certain pharmaceutical and biological features. Three QSPR models to predict the energies of long-wavelength absorption in buffers with pH=2.5 and pH=7.0, as well as in methanol, were developed and validated in this study. These models can be further used to predict the long-wavelength absorption energies of untested substances (if they are structurally similar to the training compounds). Copyright © 2014 Elsevier B.V. All rights reserved.

A novel and reliable computational intelligence system for breast cancer detection.

PubMed

Zadeh Shirazi, Amin; Seyyed Mahdavi Chabok, Seyyed Javad; Mohammadi, Zahra

2018-05-01

Cancer is the second important morbidity and mortality factor among women and the most incident type is breast cancer. This paper suggests a hybrid computational intelligence model based on unsupervised and supervised learning techniques, i.e., self-organizing map (SOM) and complex-valued neural network (CVNN), for reliable detection of breast cancer. The dataset used in this paper consists of 822 patients with five features (patient's breast mass shape, margin, density, patient's age, and Breast Imaging Reporting and Data System assessment). The proposed model was used for the first time and can be categorized in two stages. In the first stage, considering the input features, SOM technique was used to cluster the patients with the most similarity. Then, in the second stage, for each cluster, the patient's features were applied to complex-valued neural network and dealt with to classify breast cancer severity (benign or malign). The obtained results corresponding to each patient were compared to the medical diagnosis results using receiver operating characteristic analyses and confusion matrix. In the testing phase, health and disease detection ratios were 94 and 95%, respectively. Accordingly, the superiority of the proposed model was proved and can be used for reliable and robust detection of breast cancer.

Integrated, Independent and Individual Learning Activities, First and Second Grades. Summer Learning Activities, Second and Third Grades. Boston-Northampton Language Arts Program, ESEA - 1965, Projects to Advance Creativity in Education.

ERIC Educational Resources Information Center

Baldwin, Virginia

The purpose of this document is to help teachers stimulate children and provide successful learning experiences in order to develop positive self-concepts. Part I contains lists of suggestions of activities for unsupervised work at the following centers: (1) language, (2) chalk, (3) math, (4) measuring, (5) music, (6) games, toys, and puzzles, (7)…

An unsupervised machine learning model for discovering latent infectious diseases using social media data.

PubMed

Lim, Sunghoon; Tucker, Conrad S; Kumara, Soundar

2017-02-01

The authors of this work propose an unsupervised machine learning model that has the ability to identify real-world latent infectious diseases by mining social media data. In this study, a latent infectious disease is defined as a communicable disease that has not yet been formalized by national public health institutes and explicitly communicated to the general public. Most existing approaches to modeling infectious-disease-related knowledge discovery through social media networks are top-down approaches that are based on already known information, such as the names of diseases and their symptoms. In existing top-down approaches, necessary but unknown information, such as disease names and symptoms, is mostly unidentified in social media data until national public health institutes have formalized that disease. Most of the formalizing processes for latent infectious diseases are time consuming. Therefore, this study presents a bottom-up approach for latent infectious disease discovery in a given location without prior information, such as disease names and related symptoms. Social media messages with user and temporal information are extracted during the data preprocessing stage. An unsupervised sentiment analysis model is then presented. Users' expressions about symptoms, body parts, and pain locations are also identified from social media data. Then, symptom weighting vectors for each individual and time period are created, based on their sentiment and social media expressions. Finally, latent-infectious-disease-related information is retrieved from individuals' symptom weighting vectors. Twitter data from August 2012 to May 2013 are used to validate this study. Real electronic medical records for 104 individuals, who were diagnosed with influenza in the same period, are used to serve as ground truth validation. The results are promising, with the highest precision, recall, and F 1 score values of 0.773, 0.680, and 0.724, respectively. This work uses individuals' social media messages to identify latent infectious diseases, without prior information, quicker than when the disease(s) is formalized by national public health institutes. In particular, the unsupervised machine learning model using user, textual, and temporal information in social media data, along with sentiment analysis, identifies latent infectious diseases in a given location. Copyright © 2016 Elsevier Inc. All rights reserved.

An unsupervised method for quantifying the behavior of paired animals

NASA Astrophysics Data System (ADS)

Klibaite, Ugne; Berman, Gordon J.; Cande, Jessica; Stern, David L.; Shaevitz, Joshua W.

2017-02-01

Behaviors involving the interaction of multiple individuals are complex and frequently crucial for an animal’s survival. These interactions, ranging across sensory modalities, length scales, and time scales, are often subtle and difficult to characterize. Contextual effects on the frequency of behaviors become even more difficult to quantify when physical interaction between animals interferes with conventional data analysis, e.g. due to visual occlusion. We introduce a method for quantifying behavior in fruit fly interaction that combines high-throughput video acquisition and tracking of individuals with recent unsupervised methods for capturing an animal’s entire behavioral repertoire. We find behavioral differences between solitary flies and those paired with an individual of the opposite sex, identifying specific behaviors that are affected by social and spatial context. Our pipeline allows for a comprehensive description of the interaction between two individuals using unsupervised machine learning methods, and will be used to answer questions about the depth of complexity and variance in fruit fly courtship.

Alternatively Constrained Dictionary Learning For Image Superresolution.

PubMed

Lu, Xiaoqiang; Yuan, Yuan; Yan, Pingkun

2014-03-01

Dictionaries are crucial in sparse coding-based algorithm for image superresolution. Sparse coding is a typical unsupervised learning method to study the relationship between the patches of high-and low-resolution images. However, most of the sparse coding methods for image superresolution fail to simultaneously consider the geometrical structure of the dictionary and the corresponding coefficients, which may result in noticeable superresolution reconstruction artifacts. In other words, when a low-resolution image and its corresponding high-resolution image are represented in their feature spaces, the two sets of dictionaries and the obtained coefficients have intrinsic links, which has not yet been well studied. Motivated by the development on nonlocal self-similarity and manifold learning, a novel sparse coding method is reported to preserve the geometrical structure of the dictionary and the sparse coefficients of the data. Moreover, the proposed method can preserve the incoherence of dictionary entries and provide the sparse coefficients and learned dictionary from a new perspective, which have both reconstruction and discrimination properties to enhance the learning performance. Furthermore, to utilize the model of the proposed method more effectively for single-image superresolution, this paper also proposes a novel dictionary-pair learning method, which is named as two-stage dictionary training. Extensive experiments are carried out on a large set of images comparing with other popular algorithms for the same purpose, and the results clearly demonstrate the effectiveness of the proposed sparse representation model and the corresponding dictionary learning algorithm.

Visualization and unsupervised predictive clustering of high-dimensional multimodal neuroimaging data.

PubMed

Mwangi, Benson; Soares, Jair C; Hasan, Khader M

2014-10-30

Neuroimaging machine learning studies have largely utilized supervised algorithms - meaning they require both neuroimaging scan data and corresponding target variables (e.g. healthy vs. diseased) to be successfully 'trained' for a prediction task. Noticeably, this approach may not be optimal or possible when the global structure of the data is not well known and the researcher does not have an a priori model to fit the data. We set out to investigate the utility of an unsupervised machine learning technique; t-distributed stochastic neighbour embedding (t-SNE) in identifying 'unseen' sample population patterns that may exist in high-dimensional neuroimaging data. Multimodal neuroimaging scans from 92 healthy subjects were pre-processed using atlas-based methods, integrated and input into the t-SNE algorithm. Patterns and clusters discovered by the algorithm were visualized using a 2D scatter plot and further analyzed using the K-means clustering algorithm. t-SNE was evaluated against classical principal component analysis. Remarkably, based on unlabelled multimodal scan data, t-SNE separated study subjects into two very distinct clusters which corresponded to subjects' gender labels (cluster silhouette index value=0.79). The resulting clusters were used to develop an unsupervised minimum distance clustering model which identified 93.5% of subjects' gender. Notably, from a neuropsychiatric perspective this method may allow discovery of data-driven disease phenotypes or sub-types of treatment responders. Copyright © 2014 Elsevier B.V. All rights reserved.

Implementing Machine Learning in Radiology Practice and Research.

PubMed

Kohli, Marc; Prevedello, Luciano M; Filice, Ross W; Geis, J Raymond

2017-04-01

The purposes of this article are to describe concepts that radiologists should understand to evaluate machine learning projects, including common algorithms, supervised as opposed to unsupervised techniques, statistical pitfalls, and data considerations for training and evaluation, and to briefly describe ethical dilemmas and legal risk. Machine learning includes a broad class of computer programs that improve with experience. The complexity of creating, training, and monitoring machine learning indicates that the success of the algorithms will require radiologist involvement for years to come, leading to engagement rather than replacement.

Machine learning in cardiovascular medicine: are we there yet?

PubMed

Shameer, Khader; Johnson, Kipp W; Glicksberg, Benjamin S; Dudley, Joel T; Sengupta, Partho P

2018-01-19

Artificial intelligence (AI) broadly refers to analytical algorithms that iteratively learn from data, allowing computers to find hidden insights without being explicitly programmed where to look. These include a family of operations encompassing several terms like machine learning, cognitive learning, deep learning and reinforcement learning-based methods that can be used to integrate and interpret complex biomedical and healthcare data in scenarios where traditional statistical methods may not be able to perform. In this review article, we discuss the basics of machine learning algorithms and what potential data sources exist; evaluate the need for machine learning; and examine the potential limitations and challenges of implementing machine in the context of cardiovascular medicine. The most promising avenues for AI in medicine are the development of automated risk prediction algorithms which can be used to guide clinical care; use of unsupervised learning techniques to more precisely phenotype complex disease; and the implementation of reinforcement learning algorithms to intelligently augment healthcare providers. The utility of a machine learning-based predictive model will depend on factors including data heterogeneity, data depth, data breadth, nature of modelling task, choice of machine learning and feature selection algorithms, and orthogonal evidence. A critical understanding of the strength and limitations of various methods and tasks amenable to machine learning is vital. By leveraging the growing corpus of big data in medicine, we detail pathways by which machine learning may facilitate optimal development of patient-specific models for improving diagnoses, intervention and outcome in cardiovascular medicine. © Article author(s) (or their employer(s) unless otherwise stated in the text of the article) 2018. All rights reserved. No commercial use is permitted unless otherwise expressly granted.

Unsupervised consensus cluster analysis of [18F]-fluoroethyl-L-tyrosine positron emission tomography identified textural features for the diagnosis of pseudoprogression in high-grade glioma.

PubMed

Kebir, Sied; Khurshid, Zain; Gaertner, Florian C; Essler, Markus; Hattingen, Elke; Fimmers, Rolf; Scheffler, Björn; Herrlinger, Ulrich; Bundschuh, Ralph A; Glas, Martin

2017-01-31

Timely detection of pseudoprogression (PSP) is crucial for the management of patients with high-grade glioma (HGG) but remains difficult. Textural features of O-(2-[18F]fluoroethyl)-L-tyrosine positron emission tomography (FET-PET) mirror tumor uptake heterogeneity; some of them may be associated with tumor progression. Fourteen patients with HGG and suspected of PSP underwent FET-PET imaging. A set of 19 conventional and textural FET-PET features were evaluated and subjected to unsupervised consensus clustering. The final diagnosis of true progression vs. PSP was based on follow-up MRI using RANO criteria. Three robust clusters have been identified based on 10 predominantly textural FET-PET features. None of the patients with PSP fell into cluster 2, which was associated with high values for textural FET-PET markers of uptake heterogeneity. Three out of 4 patients with PSP were assigned to cluster 3 that was largely associated with low values of textural FET-PET features. By comparison, tumor-to-normal brain ratio (TNRmax) at the optimal cutoff 2.1 was less predictive of PSP (negative predictive value 57% for detecting true progression, p=0.07 vs. 75% with cluster 3, p=0.04). Clustering based on textural O-(2-[18F]fluoroethyl)-L-tyrosine PET features may provide valuable information in assessing the elusive phenomenon of pseudoprogression.

Closed-form expressions of some stochastic adapting equations for nonlinear adaptive activation function neurons.

PubMed

Fiori, Simone

2003-12-01

In recent work, we introduced nonlinear adaptive activation function (FAN) artificial neuron models, which learn their activation functions in an unsupervised way by information-theoretic adapting rules. We also applied networks of these neurons to some blind signal processing problems, such as independent component analysis and blind deconvolution. The aim of this letter is to study some fundamental aspects of FAN units' learning by investigating the properties of the associated learning differential equation systems.

Methods of Sparse Modeling and Dimensionality Reduction to Deal with Big Data

DTIC Science & Technology

2015-04-01

supervised learning (c). Our framework consists of two separate phases: (a) first find an initial space in an unsupervised manner; then (b) utilize label...model that can learn thousands of topics from a large set of documents and infer the topic mixture of each document, 2) a supervised dimension reduction...model that can learn thousands of topics from a large set of documents and infer the topic mixture of each document, (i) a method of supervised

Automated glioblastoma segmentation based on a multiparametric structured unsupervised classification.

PubMed

Juan-Albarracín, Javier; Fuster-Garcia, Elies; Manjón, José V; Robles, Montserrat; Aparici, F; Martí-Bonmatí, L; García-Gómez, Juan M

2015-01-01

Automatic brain tumour segmentation has become a key component for the future of brain tumour treatment. Currently, most of brain tumour segmentation approaches arise from the supervised learning standpoint, which requires a labelled training dataset from which to infer the models of the classes. The performance of these models is directly determined by the size and quality of the training corpus, whose retrieval becomes a tedious and time-consuming task. On the other hand, unsupervised approaches avoid these limitations but often do not reach comparable results than the supervised methods. In this sense, we propose an automated unsupervised method for brain tumour segmentation based on anatomical Magnetic Resonance (MR) images. Four unsupervised classification algorithms, grouped by their structured or non-structured condition, were evaluated within our pipeline. Considering the non-structured algorithms, we evaluated K-means, Fuzzy K-means and Gaussian Mixture Model (GMM), whereas as structured classification algorithms we evaluated Gaussian Hidden Markov Random Field (GHMRF). An automated postprocess based on a statistical approach supported by tissue probability maps is proposed to automatically identify the tumour classes after the segmentations. We evaluated our brain tumour segmentation method with the public BRAin Tumor Segmentation (BRATS) 2013 Test and Leaderboard datasets. Our approach based on the GMM model improves the results obtained by most of the supervised methods evaluated with the Leaderboard set and reaches the second position in the ranking. Our variant based on the GHMRF achieves the first position in the Test ranking of the unsupervised approaches and the seventh position in the general Test ranking, which confirms the method as a viable alternative for brain tumour segmentation.

Discovery of Deep Structure from Unlabeled Data

DTIC Science & Technology

2014-11-01

GPU processors . To evaluate the unsupervised learning component of the algorithms (which has become of less importance in the era of “big data...representations to those in biological visual, auditory, and somatosensory cortex ; and ran numerous control experiments investigating the impact of

Classification of prostate cancer grade using temporal ultrasound: in vivo feasibility study

NASA Astrophysics Data System (ADS)

Ghavidel, Sahar; Imani, Farhad; Khallaghi, Siavash; Gibson, Eli; Khojaste, Amir; Gaed, Mena; Moussa, Madeleine; Gomez, Jose A.; Siemens, D. Robert; Leveridge, Michael; Chang, Silvia; Fenster, Aaron; Ward, Aaron D.; Abolmaesumi, Purang; Mousavi, Parvin

2016-03-01

Temporal ultrasound has been shown to have high classification accuracy in differentiating cancer from benign tissue. In this paper, we extend the temporal ultrasound method to classify lower grade Prostate Cancer (PCa) from all other grades. We use a group of nine patients with mostly lower grade PCa, where cancerous regions are also limited. A critical challenge is to train a classifier with limited aggressive cancerous tissue compared to low grade cancerous tissue. To resolve the problem of imbalanced data, we use Synthetic Minority Oversampling Technique (SMOTE) to generate synthetic samples for the minority class. We calculate spectral features of temporal ultrasound data and perform feature selection using Random Forests. In leave-one-patient-out cross-validation strategy, an area under receiver operating characteristic curve (AUC) of 0.74 is achieved with overall sensitivity and specificity of 70%. Using an unsupervised learning approach prior to proposed method improves sensitivity and AUC to 80% and 0.79. This work represents promising results to classify lower and higher grade PCa with limited cancerous training samples, using temporal ultrasound.

Differentiation of Uterine Leiomyosarcoma from Atypical Leiomyoma: Diagnostic Accuracy of Qualitative MR Imaging Features and Feasibility of Texture Analysis.

PubMed

Lakhman, Yulia; Veeraraghavan, Harini; Chaim, Joshua; Feier, Diana; Goldman, Debra A; Moskowitz, Chaya S; Nougaret, Stephanie; Sosa, Ramon E; Vargas, Hebert Alberto; Soslow, Robert A; Abu-Rustum, Nadeem R; Hricak, Hedvig; Sala, Evis

2017-07-01

To investigate whether qualitative magnetic resonance (MR) features can distinguish leiomyosarcoma (LMS) from atypical leiomyoma (ALM) and assess the feasibility of texture analysis (TA). This retrospective study included 41 women (ALM = 22, LMS = 19) imaged with MRI prior to surgery. Two readers (R1, R2) evaluated each lesion for qualitative MR features. Associations between MR features and LMS were evaluated with Fisher's exact test. Accuracy measures were calculated for the four most significant features. TA was performed for 24 patients (ALM = 14, LMS = 10) with uniform imaging following lesion segmentation on axial T2-weighted images. Texture features were pre-selected using Wilcoxon signed-rank test with Bonferroni correction and analyzed with unsupervised clustering to separate LMS from ALM. Four qualitative MR features most strongly associated with LMS were nodular borders, haemorrhage, "T2 dark" area(s), and central unenhanced area(s) (p ≤ 0.0001 each feature/reader). The highest sensitivity [1.00 (95%CI:0.82-1.00)/0.95 (95%CI: 0.74-1.00)] and specificity [0.95 (95%CI:0.77-1.00)/1.00 (95%CI:0.85-1.00)] were achieved for R1/R2, respectively, when a lesion had ≥3 of these four features. Sixteen texture features differed significantly between LMS and ALM (p-values: <0.001-0.036). Unsupervised clustering achieved accuracy of 0.75 (sensitivity: 0.70; specificity: 0.79). Combination of ≥3 qualitative MR features accurately distinguished LMS from ALM. TA was feasible. • Four qualitative MR features demonstrated the strongest statistical association with LMS. • Combination of ≥3 these features could accurately differentiate LMS from ALM. • Texture analysis was a feasible semi-automated approach for lesion categorization.

Bayesian Fusion of Color and Texture Segmentations

NASA Technical Reports Server (NTRS)

Manduchi, Roberto

2000-01-01

In many applications one would like to use information from both color and texture features in order to segment an image. We propose a novel technique to combine "soft" segmentations computed for two or more features independently. Our algorithm merges models according to a mean entropy criterion, and allows to choose the appropriate number of classes for the final grouping. This technique also allows to improve the quality of supervised classification based on one feature (e.g. color) by merging information from unsupervised segmentation based on another feature (e.g., texture.)

Introduction to Concepts in Artificial Neural Networks

NASA Technical Reports Server (NTRS)

Niebur, Dagmar

1995-01-01

This introduction to artificial neural networks summarizes some basic concepts of computational neuroscience and the resulting models of artificial neurons. The terminology of biological and artificial neurons, biological and machine learning and neural processing is introduced. The concepts of supervised and unsupervised learning are explained with examples from the power system area. Finally, a taxonomy of different types of neurons and different classes of artificial neural networks is presented.

Spectral Transfer Learning Using Information Geometry for a User-Independent Brain-Computer Interface

DOE PAGES

Waytowich, Nicholas R.; Lawhern, Vernon J.; Bohannon, Addison W.; ...

2016-09-22

Recent advances in signal processing and machine learning techniques have enabled the application of Brain-Computer Interface (BCI) technologies to fields such as medicine, industry, and recreation; however, BCIs still suffer from the requirement of frequent calibration sessions due to the intra- and inter-individual variability of brain-signals, which makes calibration suppression through transfer learning an area of increasing interest for the development of practical BCI systems. In this paper, we present an unsupervised transfer method (spectral transfer using information geometry,STIG),which ranks and combines unlabeled predictions from an ensemble of information geometry classifiers built on data from individual training subjects. The STIGmore » method is validated in both off-line and real-time feedback analysis during a rapid serial visual presentation task (RSVP). For detection of single-trial, event-related potentials (ERPs), the proposed method can significantly outperform existing calibration-free techniques as well as out perform traditional within-subject calibration techniques when limited data is available. Here, this method demonstrates that unsupervised transfer learning for single-trial detection in ERP-based BCIs can be achieved without the requirement of costly training data, representing a step-forward in the overall goal of achieving a practical user-independent BCI system.« less

Spectral Transfer Learning Using Information Geometry for a User-Independent Brain-Computer Interface

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

Waytowich, Nicholas R.; Lawhern, Vernon J.; Bohannon, Addison W.

Recent advances in signal processing and machine learning techniques have enabled the application of Brain-Computer Interface (BCI) technologies to fields such as medicine, industry, and recreation; however, BCIs still suffer from the requirement of frequent calibration sessions due to the intra- and inter-individual variability of brain-signals, which makes calibration suppression through transfer learning an area of increasing interest for the development of practical BCI systems. In this paper, we present an unsupervised transfer method (spectral transfer using information geometry,STIG),which ranks and combines unlabeled predictions from an ensemble of information geometry classifiers built on data from individual training subjects. The STIGmore » method is validated in both off-line and real-time feedback analysis during a rapid serial visual presentation task (RSVP). For detection of single-trial, event-related potentials (ERPs), the proposed method can significantly outperform existing calibration-free techniques as well as out perform traditional within-subject calibration techniques when limited data is available. Here, this method demonstrates that unsupervised transfer learning for single-trial detection in ERP-based BCIs can be achieved without the requirement of costly training data, representing a step-forward in the overall goal of achieving a practical user-independent BCI system.« less

Case-based fracture image retrieval.

PubMed

Zhou, Xin; Stern, Richard; Müller, Henning

2012-05-01

Case-based fracture image retrieval can assist surgeons in decisions regarding new cases by supplying visually similar past cases. This tool may guide fracture fixation and management through comparison of long-term outcomes in similar cases. A fracture image database collected over 10 years at the orthopedic service of the University Hospitals of Geneva was used. This database contains 2,690 fracture cases associated with 43 classes (based on the AO/OTA classification). A case-based retrieval engine was developed and evaluated using retrieval precision as a performance metric. Only cases in the same class as the query case are considered as relevant. The scale-invariant feature transform (SIFT) is used for image analysis. Performance evaluation was computed in terms of mean average precision (MAP) and early precision (P10, P30). Retrieval results produced with the GNU image finding tool (GIFT) were used as a baseline. Two sampling strategies were evaluated. One used a dense 40 × 40 pixel grid sampling, and the second one used the standard SIFT features. Based on dense pixel grid sampling, three unsupervised feature selection strategies were introduced to further improve retrieval performance. With dense pixel grid sampling, the image is divided into 1,600 (40 × 40) square blocks. The goal is to emphasize the salient regions (blocks) and ignore irrelevant regions. Regions are considered as important when a high variance of the visual features is found. The first strategy is to calculate the variance of all descriptors on the global database. The second strategy is to calculate the variance of all descriptors for each case. A third strategy is to perform a thumbnail image clustering in a first step and then to calculate the variance for each cluster. Finally, a fusion between a SIFT-based system and GIFT is performed. A first comparison on the selection of sampling strategies using SIFT features shows that dense sampling using a pixel grid (MAP = 0.18) outperformed the SIFT detector-based sampling approach (MAP = 0.10). In a second step, three unsupervised feature selection strategies were evaluated. A grid parameter search is applied to optimize parameters for feature selection and clustering. Results show that using half of the regions (700 or 800) obtains the best performance for all three strategies. Increasing the number of clusters in clustering can also improve the retrieval performance. The SIFT descriptor variance in each case gave the best indication of saliency for the regions (MAP = 0.23), better than the other two strategies (MAP = 0.20 and 0.21). Combining GIFT (MAP = 0.23) and the best SIFT strategy (MAP = 0.23) produced significantly better results (MAP = 0.27) than each system alone. A case-based fracture retrieval engine was developed and is available for online demonstration. SIFT is used to extract local features, and three feature selection strategies were introduced and evaluated. A baseline using the GIFT system was used to evaluate the salient point-based approaches. Without supervised learning, SIFT-based systems with optimized parameters slightly outperformed the GIFT system. A fusion of the two approaches shows that the information contained in the two approaches is complementary. Supervised learning on the feature space is foreseen as the next step of this study.

Network Supervision of Adult Experience and Learning Dependent Sensory Cortical Plasticity.

PubMed

Blake, David T

2017-06-18

The brain is capable of remodeling throughout life. The sensory cortices provide a useful preparation for studying neuroplasticity both during development and thereafter. In adulthood, sensory cortices change in the cortical area activated by behaviorally relevant stimuli, by the strength of response within that activated area, and by the temporal profiles of those responses. Evidence supports forms of unsupervised, reinforcement, and fully supervised network learning rules. Studies on experience-dependent plasticity have mostly not controlled for learning, and they find support for unsupervised learning mechanisms. Changes occur with greatest ease in neurons containing α-CamKII, which are pyramidal neurons in layers II/III and layers V/VI. These changes use synaptic mechanisms including long term depression. Synaptic strengthening at NMDA-containing synapses does occur, but its weak association with activity suggests other factors also initiate changes. Studies that control learning find support of reinforcement learning rules and limited evidence of other forms of supervised learning. Behaviorally associating a stimulus with reinforcement leads to a strengthening of cortical response strength and enlarging of response area with poor selectivity. Associating a stimulus with omission of reinforcement leads to a selective weakening of responses. In some preparations in which these associations are not as clearly made, neurons with the most informative discharges are relatively stronger after training. Studies analyzing the temporal profile of responses associated with omission of reward, or of plasticity in studies with different discriminanda but statistically matched stimuli, support the existence of limited supervised network learning. © 2017 American Physiological Society. Compr Physiol 7:977-1008, 2017. Copyright © 2017 John Wiley & Sons, Inc.

Deep Convolutional Neural Networks for Computer-Aided Detection: CNN Architectures, Dataset Characteristics and Transfer Learning.

PubMed

Shin, Hoo-Chang; Roth, Holger R; Gao, Mingchen; Lu, Le; Xu, Ziyue; Nogues, Isabella; Yao, Jianhua; Mollura, Daniel; Summers, Ronald M

2016-05-01

Remarkable progress has been made in image recognition, primarily due to the availability of large-scale annotated datasets and deep convolutional neural networks (CNNs). CNNs enable learning data-driven, highly representative, hierarchical image features from sufficient training data. However, obtaining datasets as comprehensively annotated as ImageNet in the medical imaging domain remains a challenge. There are currently three major techniques that successfully employ CNNs to medical image classification: training the CNN from scratch, using off-the-shelf pre-trained CNN features, and conducting unsupervised CNN pre-training with supervised fine-tuning. Another effective method is transfer learning, i.e., fine-tuning CNN models pre-trained from natural image dataset to medical image tasks. In this paper, we exploit three important, but previously understudied factors of employing deep convolutional neural networks to computer-aided detection problems. We first explore and evaluate different CNN architectures. The studied models contain 5 thousand to 160 million parameters, and vary in numbers of layers. We then evaluate the influence of dataset scale and spatial image context on performance. Finally, we examine when and why transfer learning from pre-trained ImageNet (via fine-tuning) can be useful. We study two specific computer-aided detection (CADe) problems, namely thoraco-abdominal lymph node (LN) detection and interstitial lung disease (ILD) classification. We achieve the state-of-the-art performance on the mediastinal LN detection, and report the first five-fold cross-validation classification results on predicting axial CT slices with ILD categories. Our extensive empirical evaluation, CNN model analysis and valuable insights can be extended to the design of high performance CAD systems for other medical imaging tasks.

Unsupervised discovery of information structure in biomedical documents.

PubMed

Kiela, Douwe; Guo, Yufan; Stenius, Ulla; Korhonen, Anna

2015-04-01

Information structure (IS) analysis is a text mining technique, which classifies text in biomedical articles into categories that capture different types of information, such as objectives, methods, results and conclusions of research. It is a highly useful technique that can support a range of Biomedical Text Mining tasks and can help readers of biomedical literature find information of interest faster, accelerating the highly time-consuming process of literature review. Several approaches to IS analysis have been presented in the past, with promising results in real-world biomedical tasks. However, all existing approaches, even weakly supervised ones, require several hundreds of hand-annotated training sentences specific to the domain in question. Because biomedicine is subject to considerable domain variation, such annotations are expensive to obtain. This makes the application of IS analysis across biomedical domains difficult. In this article, we investigate an unsupervised approach to IS analysis and evaluate the performance of several unsupervised methods on a large corpus of biomedical abstracts collected from PubMed. Our best unsupervised algorithm (multilevel-weighted graph clustering algorithm) performs very well on the task, obtaining over 0.70 F scores for most IS categories when applied to well-known IS schemes. This level of performance is close to that of lightly supervised IS methods and has proven sufficient to aid a range of practical tasks. Thus, using an unsupervised approach, IS could be applied to support a wide range of tasks across sub-domains of biomedicine. We also demonstrate that unsupervised learning brings novel insights into IS of biomedical literature and discovers information categories that are not present in any of the existing IS schemes. The annotated corpus and software are available at http://www.cl.cam.ac.uk/∼dk427/bio14info.html. © The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.

Evaluating Mixture Modeling for Clustering: Recommendations and Cautions

ERIC Educational Resources Information Center

Steinley, Douglas; Brusco, Michael J.

2011-01-01

This article provides a large-scale investigation into several of the properties of mixture-model clustering techniques (also referred to as latent class cluster analysis, latent profile analysis, model-based clustering, probabilistic clustering, Bayesian classification, unsupervised learning, and finite mixture models; see Vermunt & Magdison,…

Self-Organizing Hidden Markov Model Map (SOHMMM).

PubMed

Ferles, Christos; Stafylopatis, Andreas

2013-12-01

A hybrid approach combining the Self-Organizing Map (SOM) and the Hidden Markov Model (HMM) is presented. The Self-Organizing Hidden Markov Model Map (SOHMMM) establishes a cross-section between the theoretic foundations and algorithmic realizations of its constituents. The respective architectures and learning methodologies are fused in an attempt to meet the increasing requirements imposed by the properties of deoxyribonucleic acid (DNA), ribonucleic acid (RNA), and protein chain molecules. The fusion and synergy of the SOM unsupervised training and the HMM dynamic programming algorithms bring forth a novel on-line gradient descent unsupervised learning algorithm, which is fully integrated into the SOHMMM. Since the SOHMMM carries out probabilistic sequence analysis with little or no prior knowledge, it can have a variety of applications in clustering, dimensionality reduction and visualization of large-scale sequence spaces, and also, in sequence discrimination, search and classification. Two series of experiments based on artificial sequence data and splice junction gene sequences demonstrate the SOHMMM's characteristics and capabilities. Copyright © 2013 Elsevier Ltd. All rights reserved.

Automatic Polyp Detection via A Novel Unified Bottom-up and Top-down Saliency Approach.

PubMed

Yuan, Yixuan; Li, Dengwang; Meng, Max Q-H

2017-07-31

In this paper, we propose a novel automatic computer-aided method to detect polyps for colonoscopy videos. To find the perceptually and semantically meaningful salient polyp regions, we first segment images into multilevel superpixels. Each level corresponds to different sizes of superpixels. Rather than adopting hand-designed features to describe these superpixels in images, we employ sparse autoencoder (SAE) to learn discriminative features in an unsupervised way. Then a novel unified bottom-up and top-down saliency method is proposed to detect polyps. In the first stage, we propose a weak bottom-up (WBU) saliency map by fusing the contrast based saliency and object-center based saliency together. The contrast based saliency map highlights image parts that show different appearances compared with surrounding areas while the object-center based saliency map emphasizes the center of the salient object. In the second stage, a strong classifier with Multiple Kernel Boosting (MKB) is learned to calculate the strong top-down (STD) saliency map based on samples directly from the obtained multi-level WBU saliency maps. We finally integrate these two stage saliency maps from all levels together to highlight polyps. Experiment results achieve 0.818 recall for saliency calculation, validating the effectiveness of our method. Extensive experiments on public polyp datasets demonstrate that the proposed saliency algorithm performs favorably against state-of-the-art saliency methods to detect polyps.

Natural image sequences constrain dynamic receptive fields and imply a sparse code.

PubMed

Häusler, Chris; Susemihl, Alex; Nawrot, Martin P

2013-11-06

In their natural environment, animals experience a complex and dynamic visual scenery. Under such natural stimulus conditions, neurons in the visual cortex employ a spatially and temporally sparse code. For the input scenario of natural still images, previous work demonstrated that unsupervised feature learning combined with the constraint of sparse coding can predict physiologically measured receptive fields of simple cells in the primary visual cortex. This convincingly indicated that the mammalian visual system is adapted to the natural spatial input statistics. Here, we extend this approach to the time domain in order to predict dynamic receptive fields that can account for both spatial and temporal sparse activation in biological neurons. We rely on temporal restricted Boltzmann machines and suggest a novel temporal autoencoding training procedure. When tested on a dynamic multi-variate benchmark dataset this method outperformed existing models of this class. Learning features on a large dataset of natural movies allowed us to model spatio-temporal receptive fields for single neurons. They resemble temporally smooth transformations of previously obtained static receptive fields and are thus consistent with existing theories. A neuronal spike response model demonstrates how the dynamic receptive field facilitates temporal and population sparseness. We discuss the potential mechanisms and benefits of a spatially and temporally sparse representation of natural visual input. Copyright © 2013 The Authors. Published by Elsevier B.V. All rights reserved.

Hard exudates segmentation based on learned initial seeds and iterative graph cut.

PubMed

Kusakunniran, Worapan; Wu, Qiang; Ritthipravat, Panrasee; Zhang, Jian

2018-05-01

(Background and Objective): The occurrence of hard exudates is one of the early signs of diabetic retinopathy which is one of the leading causes of the blindness. Many patients with diabetic retinopathy lose their vision because of the late detection of the disease. Thus, this paper is to propose a novel method of hard exudates segmentation in retinal images in an automatic way. (Methods): The existing methods are based on either supervised or unsupervised learning techniques. In addition, the learned segmentation models may often cause miss-detection and/or fault-detection of hard exudates, due to the lack of rich characteristics, the intra-variations, and the similarity with other components in the retinal image. Thus, in this paper, the supervised learning based on the multilayer perceptron (MLP) is only used to identify initial seeds with high confidences to be hard exudates. Then, the segmentation is finalized by unsupervised learning based on the iterative graph cut (GC) using clusters of initial seeds. Also, in order to reduce color intra-variations of hard exudates in different retinal images, the color transfer (CT) is applied to normalize their color information, in the pre-processing step. (Results): The experiments and comparisons with the other existing methods are based on the two well-known datasets, e_ophtha EX and DIARETDB1. It can be seen that the proposed method outperforms the other existing methods in the literature, with the sensitivity in the pixel-level of 0.891 for the DIARETDB1 dataset and 0.564 for the e_ophtha EX dataset. The cross datasets validation where the training process is performed on one dataset and the testing process is performed on another dataset is also evaluated in this paper, in order to illustrate the robustness of the proposed method. (Conclusions): This newly proposed method integrates the supervised learning and unsupervised learning based techniques. It achieves the improved performance, when compared with the existing methods in the literature. The robustness of the proposed method for the scenario of cross datasets could enhance its practical usage. That is, the trained model could be more practical for unseen data in the real-world situation, especially when the capturing environments of training and testing images are not the same. Copyright © 2018 Elsevier B.V. All rights reserved.

Adaptive learning in a compartmental model of visual cortex—how feedback enables stable category learning and refinement

PubMed Central

Layher, Georg; Schrodt, Fabian; Butz, Martin V.; Neumann, Heiko

2014-01-01

The categorization of real world objects is often reflected in the similarity of their visual appearances. Such categories of objects do not necessarily form disjunct sets of objects, neither semantically nor visually. The relationship between categories can often be described in terms of a hierarchical structure. For instance, tigers and leopards build two separate mammalian categories, both of which are subcategories of the category Felidae. In the last decades, the unsupervised learning of categories of visual input stimuli has been addressed by numerous approaches in machine learning as well as in computational neuroscience. However, the question of what kind of mechanisms might be involved in the process of subcategory learning, or category refinement, remains a topic of active investigation. We propose a recurrent computational network architecture for the unsupervised learning of categorial and subcategorial visual input representations. During learning, the connection strengths of bottom-up weights from input to higher-level category representations are adapted according to the input activity distribution. In a similar manner, top-down weights learn to encode the characteristics of a specific stimulus category. Feedforward and feedback learning in combination realize an associative memory mechanism, enabling the selective top-down propagation of a category's feedback weight distribution. We suggest that the difference between the expected input encoded in the projective field of a category node and the current input pattern controls the amplification of feedforward-driven representations. Large enough differences trigger the recruitment of new representational resources and the establishment of additional (sub-) category representations. We demonstrate the temporal evolution of such learning and show how the proposed combination of an associative memory with a modulatory feedback integration successfully establishes category and subcategory representations. PMID:25538637

What Makes You Tick? An Empirical Study of Space Science Related Social Media Communications Using Machine Learning

NASA Astrophysics Data System (ADS)

Hwong, Y. L.; Oliver, C.; Van Kranendonk, M. J.

2016-12-01

The rise of social media has transformed the way the public engages with scientists and science organisations. `Retweet', `Like', `Share' and `Comment' are a few ways users engage with messages on Twitter and Facebook, two of the most popular social media platforms. Despite the availability of big data from these digital footprints, research into social media science communication is scant. This paper presents the results of an empirical study into the processes and outcomes of space science related social media communications using machine learning. The study is divided into two main parts. The first part is dedicated to the use of supervised learning methods to investigate the features of highly engaging messages., e.g. highly retweeted tweets and shared Facebook posts. It is hypothesised that these messages contain certain psycholinguistic features that are unique to the field of space science. We built a predictive model to forecast the engagement levels of social media posts. By using four feature sets (n-grams, psycholinguistics, grammar and social media), we were able to achieve prediction accuracies in the vicinity of 90% using three supervised learning algorithms (Naive Bayes, linear classifier and decision tree). We conducted the same experiments on social media messages from three other fields (politics, business and non-profit) and discovered several features that are exclusive to space science communications: anger, authenticity, hashtags, visual descriptions and a tentative tone. The second part of the study focuses on the extraction of topics from a corpus of texts using topic modelling. This part of the study is exploratory in nature and uses an unsupervised method called Latent Dirichlet Allocation (LDA) to uncover previously unknown topics within a large body of documents. Preliminary results indicate a strong potential of topic model algorithms to automatically uncover themes hidden within social media chatters on space related issues, with keywords such as `exoplanet', `water' and `life' being clustered together forming a topic (i.e. 'Astrobiology'). Results also demonstrate the freewheeling nature of social media conversations, while providing evidence for the role of these platforms in facilitating meaningful exchanges among science audience.

Unsupervised Sequential Outlier Detection With Deep Architectures.

PubMed

Lu, Weining; Cheng, Yu; Xiao, Cao; Chang, Shiyu; Huang, Shuai; Liang, Bin; Huang, Thomas

2017-09-01

Unsupervised outlier detection is a vital task and has high impact on a wide variety of applications domains, such as image analysis and video surveillance. It also gains long-standing attentions and has been extensively studied in multiple research areas. Detecting and taking action on outliers as quickly as possible are imperative in order to protect network and related stakeholders or to maintain the reliability of critical systems. However, outlier detection is difficult due to the one class nature and challenges in feature construction. Sequential anomaly detection is even harder with more challenges from temporal correlation in data, as well as the presence of noise and high dimensionality. In this paper, we introduce a novel deep structured framework to solve the challenging sequential outlier detection problem. We use autoencoder models to capture the intrinsic difference between outliers and normal instances and integrate the models to recurrent neural networks that allow the learning to make use of previous context as well as make the learners more robust to warp along the time axis. Furthermore, we propose to use a layerwise training procedure, which significantly simplifies the training procedure and hence helps achieve efficient and scalable training. In addition, we investigate a fine-tuning step to update all parameters set by incorporating the temporal correlation in the sequence. We further apply our proposed models to conduct systematic experiments on five real-world benchmark data sets. Experimental results demonstrate the effectiveness of our model, compared with other state-of-the-art approaches.

LCC: Light Curves Classifier

NASA Astrophysics Data System (ADS)

Vo, Martin

2017-08-01

Light Curves Classifier uses data mining and machine learning to obtain and classify desired objects. This task can be accomplished by attributes of light curves or any time series, including shapes, histograms, or variograms, or by other available information about the inspected objects, such as color indices, temperatures, and abundances. After specifying features which describe the objects to be searched, the software trains on a given training sample, and can then be used for unsupervised clustering for visualizing the natural separation of the sample. The package can be also used for automatic tuning parameters of used methods (for example, number of hidden neurons or binning ratio). Trained classifiers can be used for filtering outputs from astronomical databases or data stored locally. The Light Curve Classifier can also be used for simple downloading of light curves and all available information of queried stars. It natively can connect to OgleII, OgleIII, ASAS, CoRoT, Kepler, Catalina and MACHO, and new connectors or descriptors can be implemented. In addition to direct usage of the package and command line UI, the program can be used through a web interface. Users can create jobs for ”training” methods on given objects, querying databases and filtering outputs by trained filters. Preimplemented descriptors, classifier and connectors can be picked by simple clicks and their parameters can be tuned by giving ranges of these values. All combinations are then calculated and the best one is used for creating the filter. Natural separation of the data can be visualized by unsupervised clustering.

Learning Microbial Community Structures with Supervised and Unsupervised Non-negative Matrix Factorization.

PubMed

Cai, Yun; Gu, Hong; Kenney, Toby

2017-08-31

Learning the structure of microbial communities is critical in understanding the different community structures and functions of microbes in distinct individuals. We view microbial communities as consisting of many subcommunities which are formed by certain groups of microbes functionally dependent on each other. The focus of this paper is on methods for extracting the subcommunities from the data, in particular Non-Negative Matrix Factorization (NMF). Our methods can be applied to both OTU data and functional metagenomic data. We apply the existing unsupervised NMF method and also develop a new supervised NMF method for extracting interpretable information from classification problems. The relevance of the subcommunities identified by NMF is demonstrated by their excellent performance for classification. Through three data examples, we demonstrate how to interpret the features identified by NMF to draw meaningful biological conclusions and discover hitherto unidentified patterns in the data. Comparing whole metagenomes of various mammals, (Muegge et al., Science 332:970-974, 2011), the biosynthesis of macrolides pathway is found in hindgut-fermenting herbivores, but not carnivores. This is consistent with results in veterinary science that macrolides should not be given to non-ruminant herbivores. For time series microbiome data from various body sites (Caporaso et al., Genome Biol 12:50, 2011), a shift in the microbial communities is identified for one individual. The shift occurs at around the same time in the tongue and gut microbiomes, indicating that the shift is a genuine biological trait, rather than an artefact of the method. For whole metagenome data from IBD patients and healthy controls (Qin et al., Nature 464:59-65, 2010), we identify differences in a number of pathways (some known, others new). NMF is a powerful tool for identifying the key features of microbial communities. These identified features can not only be used to perform difficult classification problems with a high degree of accuracy, they are also very interpretable and can lead to important biological insights into the structure of the communities. In addition, NMF is a dimension-reduction method (similar to PCA) in that it reduces the extremely complex microbial data into a low-dimensional representation, allowing a number of analyses to be performed more easily-for example, searching for temporal patterns in the microbiome. When we are interested in the differences between the structures of two groups of communities, supervised NMF provides a better way to do this, while retaining all the advantages of NMF-e.g. interpretability and a simple biological intuition.

Generalising better: Applying deep learning to integrate deleteriousness prediction scores for whole-exome SNV studies.

PubMed

Korvigo, Ilia; Afanasyev, Andrey; Romashchenko, Nikolay; Skoblov, Mikhail

2018-01-01

Many automatic classifiers were introduced to aid inference of phenotypical effects of uncategorised nsSNVs (nonsynonymous Single Nucleotide Variations) in theoretical and medical applications. Lately, several meta-estimators have been proposed that combine different predictors, such as PolyPhen and SIFT, to integrate more information in a single score. Although many advances have been made in feature design and machine learning algorithms used, the shortage of high-quality reference data along with the bias towards intensively studied in vitro models call for improved generalisation ability in order to further increase classification accuracy and handle records with insufficient data. Since a meta-estimator basically combines different scoring systems with highly complicated nonlinear relationships, we investigated how deep learning (supervised and unsupervised), which is particularly efficient at discovering hierarchies of features, can improve classification performance. While it is believed that one should only use deep learning for high-dimensional input spaces and other models (logistic regression, support vector machines, Bayesian classifiers, etc) for simpler inputs, we still believe that the ability of neural networks to discover intricate structure in highly heterogenous datasets can aid a meta-estimator. We compare the performance with various popular predictors, many of which are recommended by the American College of Medical Genetics and Genomics (ACMG), as well as available deep learning-based predictors. Thanks to hardware acceleration we were able to use a computationally expensive genetic algorithm to stochastically optimise hyper-parameters over many generations. Overfitting was hindered by noise injection and dropout, limiting coadaptation of hidden units. Although we stress that this work was not conceived as a tool comparison, but rather an exploration of the possibilities of deep learning application in ensemble scores, our results show that even relatively simple modern neural networks can significantly improve both prediction accuracy and coverage. We provide open-access to our finest model via the web-site: http://score.generesearch.ru/services/badmut/.

Sequential visibility-graph motifs

NASA Astrophysics Data System (ADS)

Iacovacci, Jacopo; Lacasa, Lucas

2016-04-01

Visibility algorithms transform time series into graphs and encode dynamical information in their topology, paving the way for graph-theoretical time series analysis as well as building a bridge between nonlinear dynamics and network science. In this work we introduce and study the concept of sequential visibility-graph motifs, smaller substructures of n consecutive nodes that appear with characteristic frequencies. We develop a theory to compute in an exact way the motif profiles associated with general classes of deterministic and stochastic dynamics. We find that this simple property is indeed a highly informative and computationally efficient feature capable of distinguishing among different dynamics and robust against noise contamination. We finally confirm that it can be used in practice to perform unsupervised learning, by extracting motif profiles from experimental heart-rate series and being able, accordingly, to disentangle meditative from other relaxation states. Applications of this general theory include the automatic classification and description of physical, biological, and financial time series.

Intrusion Detection System Using Deep Neural Network for In-Vehicle Network Security.

PubMed

Kang, Min-Joo; Kang, Je-Won

2016-01-01

A novel intrusion detection system (IDS) using a deep neural network (DNN) is proposed to enhance the security of in-vehicular network. The parameters building the DNN structure are trained with probability-based feature vectors that are extracted from the in-vehicular network packets. For a given packet, the DNN provides the probability of each class discriminating normal and attack packets, and, thus the sensor can identify any malicious attack to the vehicle. As compared to the traditional artificial neural network applied to the IDS, the proposed technique adopts recent advances in deep learning studies such as initializing the parameters through the unsupervised pre-training of deep belief networks (DBN), therefore improving the detection accuracy. It is demonstrated with experimental results that the proposed technique can provide a real-time response to the attack with a significantly improved detection ratio in controller area network (CAN) bus.

Intrusion Detection System Using Deep Neural Network for In-Vehicle Network Security

PubMed Central

Kang, Min-Joo

2016-01-01

A novel intrusion detection system (IDS) using a deep neural network (DNN) is proposed to enhance the security of in-vehicular network. The parameters building the DNN structure are trained with probability-based feature vectors that are extracted from the in-vehicular network packets. For a given packet, the DNN provides the probability of each class discriminating normal and attack packets, and, thus the sensor can identify any malicious attack to the vehicle. As compared to the traditional artificial neural network applied to the IDS, the proposed technique adopts recent advances in deep learning studies such as initializing the parameters through the unsupervised pre-training of deep belief networks (DBN), therefore improving the detection accuracy. It is demonstrated with experimental results that the proposed technique can provide a real-time response to the attack with a significantly improved detection ratio in controller area network (CAN) bus. PMID:27271802

Unsupervised learning of structure in spectroscopic cubes

NASA Astrophysics Data System (ADS)

Araya, M.; Mendoza, M.; Solar, M.; Mardones, D.; Bayo, A.

2018-07-01

We consider the problem of analyzing the structure of spectroscopic cubes using unsupervised machine learning techniques. We propose representing the target's signal as a homogeneous set of volumes through an iterative algorithm that separates the structured emission from the background while not overestimating the flux. Besides verifying some basic theoretical properties, the algorithm is designed to be tuned by domain experts, because its parameters have meaningful values in the astronomical context. Nevertheless, we propose a heuristic to automatically estimate the signal-to-noise ratio parameter of the algorithm directly from data. The resulting light-weighted set of samples (≤ 1% compared to the original data) offer several advantages. For instance, it is statistically correct and computationally inexpensive to apply well-established techniques of the pattern recognition and machine learning domains; such as clustering and dimensionality reduction algorithms. We use ALMA science verification data to validate our method, and present examples of the operations that can be performed by using the proposed representation. Even though this approach is focused on providing faster and better analysis tools for the end-user astronomer, it also opens the possibility of content-aware data discovery by applying our algorithm to big data.

Data Mining for Anomaly Detection

NASA Technical Reports Server (NTRS)

Biswas, Gautam; Mack, Daniel; Mylaraswamy, Dinkar; Bharadwaj, Raj

2013-01-01

The Vehicle Integrated Prognostics Reasoner (VIPR) program describes methods for enhanced diagnostics as well as a prognostic extension to current state of art Aircraft Diagnostic and Maintenance System (ADMS). VIPR introduced a new anomaly detection function for discovering previously undetected and undocumented situations, where there are clear deviations from nominal behavior. Once a baseline (nominal model of operations) is established, the detection and analysis is split between on-aircraft outlier generation and off-aircraft expert analysis to characterize and classify events that may not have been anticipated by individual system providers. Offline expert analysis is supported by data curation and data mining algorithms that can be applied in the contexts of supervised learning methods and unsupervised learning. In this report, we discuss efficient methods to implement the Kolmogorov complexity measure using compression algorithms, and run a systematic empirical analysis to determine the best compression measure. Our experiments established that the combination of the DZIP compression algorithm and CiDM distance measure provides the best results for capturing relevant properties of time series data encountered in aircraft operations. This combination was used as the basis for developing an unsupervised learning algorithm to define "nominal" flight segments using historical flight segments.

Unsupervised symmetrical trademark image retrieval in soccer telecast using wavelet energy and quadtree decomposition

NASA Astrophysics Data System (ADS)

Ong, Swee Khai; Lim, Wee Keong; Soo, Wooi King

2013-04-01

Trademark, a distinctive symbol, is used to distinguish products or services provided by a particular person, group or organization from other similar entries. As trademark represents the reputation and credit standing of the owner, it is important to differentiate one trademark from another. Many methods have been proposed to identify, classify and retrieve trademarks. However, most methods required features database and sample sets for training prior to recognition and retrieval process. In this paper, a new feature on wavelet coefficients, the localized wavelet energy, is introduced to extract features of trademarks. With this, unsupervised content-based symmetrical trademark image retrieval is proposed without the database and prior training set. The feature analysis is done by an integration of the proposed localized wavelet energy and quadtree decomposed regional symmetrical vector. The proposed framework eradicates the dependence on query database and human participation during the retrieval process. In this paper, trademarks for soccer games sponsors are the intended trademark category. Video frames from soccer telecast are extracted and processed for this study. Reasonably good localization and retrieval results on certain categories of trademarks are achieved. A distinctive symbol is used to distinguish products or services provided by a particular person, group or organization from other similar entries.

Rough-Fuzzy Clustering and Unsupervised Feature Selection for Wavelet Based MR Image Segmentation

PubMed Central

Maji, Pradipta; Roy, Shaswati

2015-01-01

Image segmentation is an indispensable process in the visualization of human tissues, particularly during clinical analysis of brain magnetic resonance (MR) images. For many human experts, manual segmentation is a difficult and time consuming task, which makes an automated brain MR image segmentation method desirable. In this regard, this paper presents a new segmentation method for brain MR images, integrating judiciously the merits of rough-fuzzy computing and multiresolution image analysis technique. The proposed method assumes that the major brain tissues, namely, gray matter, white matter, and cerebrospinal fluid from the MR images are considered to have different textural properties. The dyadic wavelet analysis is used to extract the scale-space feature vector for each pixel, while the rough-fuzzy clustering is used to address the uncertainty problem of brain MR image segmentation. An unsupervised feature selection method is introduced, based on maximum relevance-maximum significance criterion, to select relevant and significant textural features for segmentation problem, while the mathematical morphology based skull stripping preprocessing step is proposed to remove the non-cerebral tissues like skull. The performance of the proposed method, along with a comparison with related approaches, is demonstrated on a set of synthetic and real brain MR images using standard validity indices. PMID:25848961

Understanding Student Language: An Unsupervised Dialogue Act Classification Approach

ERIC Educational Resources Information Center

Ezen-Can, Aysu; Boyer, Kristy Elizabeth

2015-01-01

Within the landscape of educational data, textual natural language is an increasingly vast source of learning-centered interactions. In natural language dialogue, student contributions hold important information about knowledge and goals. Automatically modeling the dialogue act of these student utterances is crucial for scaling natural language…

76 FR 16521 - National Poison Prevention Week, 2011

Federal Register 2010, 2011, 2012, 2013, 2014

2011-03-23

..., and learn how to respond if a poison emergency should occur. Children are particularly susceptible to unintentional poisoning. More than half of all reported poison exposures involve children under the age of six, and many occur when unsupervised children find and consume medicines or harmful chemicals...

The application of artificial neural networks in astronomy

NASA Astrophysics Data System (ADS)

Li, Li-Li; Zhang, Yan-Xia; Zhao, Yong-Heng; Yang, Da-Wei

2006-12-01

Artificial Neural Networks (ANNs) are computer algorithms inspired from simple models of human central nervous system activity. They can be roughly divided into two main kinds: supervised and unsupervised. The supervised approach lays the stress on "teaching" a machine to do the work of a mention human expert, usually by showing examples for which the true answer is supplied by the expert. The unsupervised one is aimed at learning new things from the data, and most useful when the data cannot easily be plotted in a two or three dimensional space. ANNs have been used widely and successfully in various fields, for instance, pattern recognition, financial analysis, biology, engineering and so on, because they have many merits such as self-learning, self-adapting, good robustness and dynamically rapid response as well as strong capability of dealing with non-linear problems. In the last few years there has been an increasing interest toward the astronomical applications of ANNs. In this paper, the authors firstly introduce the fundamental principle of ANNs together with the architecture of the network and outline various kinds of learning algorithms and network toplogies. The specific aspects of the applications of ANNs in astronomical problems are also listed, which contain the strong capabilities of approximating to arbitrary accuracy, any nonlinear functional mapping, parallel and distributed storage, tolerance of faulty and generalization of results. They summarize the advantages and disadvantages of main ANN models available to the astronomical community. Furthermore, the application cases of ANNs in astronomy are mainly described in detail. Here, the focus is on some of the most interesting fields of its application, for example: object detection, star/galaxy classification, spectral classification, galaxy morphology classification, the estimation of photometric redshifts of galaxies and time series analysis. In addition, other kinds of applications have been only touched upon. Finally, the development and application prospects of ANNs is discussed. With the increase of quantity and the distributing complexity of astronomical data, its scientific exploitation requires a variety of automated tools, which are capable to perform huge amount of work, such as data preprocessing, feature selection, data reduction, data mining amd data analysis. ANNs, one of intelligent tools, will show more and more superiorities.

Predicting multicellular function through multi-layer tissue networks

PubMed Central

Zitnik, Marinka; Leskovec, Jure

2017-01-01

Abstract Motivation: Understanding functions of proteins in specific human tissues is essential for insights into disease diagnostics and therapeutics, yet prediction of tissue-specific cellular function remains a critical challenge for biomedicine. Results: Here, we present OhmNet, a hierarchy-aware unsupervised node feature learning approach for multi-layer networks. We build a multi-layer network, where each layer represents molecular interactions in a different human tissue. OhmNet then automatically learns a mapping of proteins, represented as nodes, to a neural embedding-based low-dimensional space of features. OhmNet encourages sharing of similar features among proteins with similar network neighborhoods and among proteins activated in similar tissues. The algorithm generalizes prior work, which generally ignores relationships between tissues, by modeling tissue organization with a rich multiscale tissue hierarchy. We use OhmNet to study multicellular function in a multi-layer protein interaction network of 107 human tissues. In 48 tissues with known tissue-specific cellular functions, OhmNet provides more accurate predictions of cellular function than alternative approaches, and also generates more accurate hypotheses about tissue-specific protein actions. We show that taking into account the tissue hierarchy leads to improved predictive power. Remarkably, we also demonstrate that it is possible to leverage the tissue hierarchy in order to effectively transfer cellular functions to a functionally uncharacterized tissue. Overall, OhmNet moves from flat networks to multiscale models able to predict a range of phenotypes spanning cellular subsystems. Availability and implementation: Source code and datasets are available at http://snap.stanford.edu/ohmnet. Contact: jure@cs.stanford.edu PMID:28881986

Automated Glioblastoma Segmentation Based on a Multiparametric Structured Unsupervised Classification

PubMed Central

Juan-Albarracín, Javier; Fuster-Garcia, Elies; Manjón, José V.; Robles, Montserrat; Aparici, F.; Martí-Bonmatí, L.; García-Gómez, Juan M.

2015-01-01

Automatic brain tumour segmentation has become a key component for the future of brain tumour treatment. Currently, most of brain tumour segmentation approaches arise from the supervised learning standpoint, which requires a labelled training dataset from which to infer the models of the classes. The performance of these models is directly determined by the size and quality of the training corpus, whose retrieval becomes a tedious and time-consuming task. On the other hand, unsupervised approaches avoid these limitations but often do not reach comparable results than the supervised methods. In this sense, we propose an automated unsupervised method for brain tumour segmentation based on anatomical Magnetic Resonance (MR) images. Four unsupervised classification algorithms, grouped by their structured or non-structured condition, were evaluated within our pipeline. Considering the non-structured algorithms, we evaluated K-means, Fuzzy K-means and Gaussian Mixture Model (GMM), whereas as structured classification algorithms we evaluated Gaussian Hidden Markov Random Field (GHMRF). An automated postprocess based on a statistical approach supported by tissue probability maps is proposed to automatically identify the tumour classes after the segmentations. We evaluated our brain tumour segmentation method with the public BRAin Tumor Segmentation (BRATS) 2013 Test and Leaderboard datasets. Our approach based on the GMM model improves the results obtained by most of the supervised methods evaluated with the Leaderboard set and reaches the second position in the ranking. Our variant based on the GHMRF achieves the first position in the Test ranking of the unsupervised approaches and the seventh position in the general Test ranking, which confirms the method as a viable alternative for brain tumour segmentation. PMID:25978453

BlobContours: adapting Blobworld for supervised color- and texture-based image segmentation

NASA Astrophysics Data System (ADS)

Vogel, Thomas; Nguyen, Dinh Quyen; Dittmann, Jana

2006-01-01

Extracting features is the first and one of the most crucial steps in recent image retrieval process. While the color features and the texture features of digital images can be extracted rather easily, the shape features and the layout features depend on reliable image segmentation. Unsupervised image segmentation, often used in image analysis, works on merely syntactical basis. That is, what an unsupervised segmentation algorithm can segment is only regions, but not objects. To obtain high-level objects, which is desirable in image retrieval, human assistance is needed. Supervised image segmentations schemes can improve the reliability of segmentation and segmentation refinement. In this paper we propose a novel interactive image segmentation technique that combines the reliability of a human expert with the precision of automated image segmentation. The iterative procedure can be considered a variation on the Blobworld algorithm introduced by Carson et al. from EECS Department, University of California, Berkeley. Starting with an initial segmentation as provided by the Blobworld framework, our algorithm, namely BlobContours, gradually updates it by recalculating every blob, based on the original features and the updated number of Gaussians. Since the original algorithm has hardly been designed for interactive processing we had to consider additional requirements for realizing a supervised segmentation scheme on the basis of Blobworld. Increasing transparency of the algorithm by applying usercontrolled iterative segmentation, providing different types of visualization for displaying the segmented image and decreasing computational time of segmentation are three major requirements which are discussed in detail.

Unsupervised consensus cluster analysis of [18F]-fluoroethyl-L-tyrosine positron emission tomography identified textural features for the diagnosis of pseudoprogression in high-grade glioma

PubMed Central

Kebir, Sied; Khurshid, Zain; Gaertner, Florian C.; Essler, Markus; Hattingen, Elke; Fimmers, Rolf; Scheffler, Björn; Herrlinger, Ulrich; Bundschuh, Ralph A.; Glas, Martin

2017-01-01

Rationale Timely detection of pseudoprogression (PSP) is crucial for the management of patients with high-grade glioma (HGG) but remains difficult. Textural features of O-(2-[18F]fluoroethyl)-L-tyrosine positron emission tomography (FET-PET) mirror tumor uptake heterogeneity; some of them may be associated with tumor progression. Methods Fourteen patients with HGG and suspected of PSP underwent FET-PET imaging. A set of 19 conventional and textural FET-PET features were evaluated and subjected to unsupervised consensus clustering. The final diagnosis of true progression vs. PSP was based on follow-up MRI using RANO criteria. Results Three robust clusters have been identified based on 10 predominantly textural FET-PET features. None of the patients with PSP fell into cluster 2, which was associated with high values for textural FET-PET markers of uptake heterogeneity. Three out of 4 patients with PSP were assigned to cluster 3 that was largely associated with low values of textural FET-PET features. By comparison, tumor-to-normal brain ratio (TNRmax) at the optimal cutoff 2.1 was less predictive of PSP (negative predictive value 57% for detecting true progression, p=0.07 vs. 75% with cluster 3, p=0.04). Principal Conclusions Clustering based on textural O-(2-[18F]fluoroethyl)-L-tyrosine PET features may provide valuable information in assessing the elusive phenomenon of pseudoprogression. PMID:28030820

An improved clustering algorithm based on reverse learning in intelligent transportation

NASA Astrophysics Data System (ADS)

Qiu, Guoqing; Kou, Qianqian; Niu, Ting

2017-05-01

With the development of artificial intelligence and data mining technology, big data has gradually entered people's field of vision. In the process of dealing with large data, clustering is an important processing method. By introducing the reverse learning method in the clustering process of PAM clustering algorithm, to further improve the limitations of one-time clustering in unsupervised clustering learning, and increase the diversity of clustering clusters, so as to improve the quality of clustering. The algorithm analysis and experimental results show that the algorithm is feasible.

3D Visualization of Machine Learning Algorithms with Astronomical Data

NASA Astrophysics Data System (ADS)

Kent, Brian R.

2016-01-01

We present innovative machine learning (ML) methods using unsupervised clustering with minimum spanning trees (MSTs) to study 3D astronomical catalogs. Utilizing Python code to build trees based on galaxy catalogs, we can render the results with the visualization suite Blender to produce interactive 360 degree panoramic videos. The catalogs and their ML results can be explored in a 3D space using mobile devices, tablets or desktop browsers. We compare the statistics of the MST results to a number of machine learning methods relating to optimization and efficiency.

Non-accidental properties, metric invariance, and encoding by neurons in a model of ventral stream visual object recognition, VisNet.

PubMed

Rolls, Edmund T; Mills, W Patrick C

2018-05-01

When objects transform into different views, some properties are maintained, such as whether the edges are convex or concave, and these non-accidental properties are likely to be important in view-invariant object recognition. The metric properties, such as the degree of curvature, may change with different views, and are less likely to be useful in object recognition. It is shown that in a model of invariant visual object recognition in the ventral visual stream, VisNet, non-accidental properties are encoded much more than metric properties by neurons. Moreover, it is shown how with the temporal trace rule training in VisNet, non-accidental properties of objects become encoded by neurons, and how metric properties are treated invariantly. We also show how VisNet can generalize between different objects if they have the same non-accidental property, because the metric properties are likely to overlap. VisNet is a 4-layer unsupervised model of visual object recognition trained by competitive learning that utilizes a temporal trace learning rule to implement the learning of invariance using views that occur close together in time. A second crucial property of this model of object recognition is, when neurons in the level corresponding to the inferior temporal visual cortex respond selectively to objects, whether neurons in the intermediate layers can respond to combinations of features that may be parts of two or more objects. In an investigation using the four sides of a square presented in every possible combination, it was shown that even though different layer 4 neurons are tuned to encode each feature or feature combination orthogonally, neurons in the intermediate layers can respond to features or feature combinations present is several objects. This property is an important part of the way in which high capacity can be achieved in the four-layer ventral visual cortical pathway. These findings concerning non-accidental properties and the use of neurons in intermediate layers of the hierarchy help to emphasise fundamental underlying principles of the computations that may be implemented in the ventral cortical visual stream used in object recognition. Copyright © 2018 Elsevier Inc. All rights reserved.

Multiview alignment hashing for efficient image search.

PubMed

Liu, Li; Yu, Mengyang; Shao, Ling

2015-03-01

Hashing is a popular and efficient method for nearest neighbor search in large-scale data spaces by embedding high-dimensional feature descriptors into a similarity preserving Hamming space with a low dimension. For most hashing methods, the performance of retrieval heavily depends on the choice of the high-dimensional feature descriptor. Furthermore, a single type of feature cannot be descriptive enough for different images when it is used for hashing. Thus, how to combine multiple representations for learning effective hashing functions is an imminent task. In this paper, we present a novel unsupervised multiview alignment hashing approach based on regularized kernel nonnegative matrix factorization, which can find a compact representation uncovering the hidden semantics and simultaneously respecting the joint probability distribution of data. In particular, we aim to seek a matrix factorization to effectively fuse the multiple information sources meanwhile discarding the feature redundancy. Since the raised problem is regarded as nonconvex and discrete, our objective function is then optimized via an alternate way with relaxation and converges to a locally optimal solution. After finding the low-dimensional representation, the hashing functions are finally obtained through multivariable logistic regression. The proposed method is systematically evaluated on three data sets: 1) Caltech-256; 2) CIFAR-10; and 3) CIFAR-20, and the results show that our method significantly outperforms the state-of-the-art multiview hashing techniques.

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.

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

Tensor decomposition-based and principal-component-analysis-based unsupervised feature extraction applied to the gene expression and methylation profiles in the brains of social insects with multiple castes.

PubMed

Taguchi, Y-H

2018-05-08

Even though coexistence of multiple phenotypes sharing the same genomic background is interesting, it remains incompletely understood. Epigenomic profiles may represent key factors, with unknown contributions to the development of multiple phenotypes, and social-insect castes are a good model for elucidation of the underlying mechanisms. Nonetheless, previous studies have failed to identify genes associated with aberrant gene expression and methylation profiles because of the lack of suitable methodology that can address this problem properly. A recently proposed principal component analysis (PCA)-based and tensor decomposition (TD)-based unsupervised feature extraction (FE) can solve this problem because these two approaches can deal with gene expression and methylation profiles even when a small number of samples is available. PCA-based and TD-based unsupervised FE methods were applied to the analysis of gene expression and methylation profiles in the brains of two social insects, Polistes canadensis and Dinoponera quadriceps. Genes associated with differential expression and methylation between castes were identified, and analysis of enrichment of Gene Ontology terms confirmed reliability of the obtained sets of genes from the biological standpoint. Biologically relevant genes, shown to be associated with significant differential gene expression and methylation between castes, were identified here for the first time. The identification of these genes may help understand the mechanisms underlying epigenetic control of development of multiple phenotypes under the same genomic conditions.

Comparing supervised learning techniques on the task of physical activity recognition.

PubMed

Dalton, A; OLaighin, G

2013-01-01

The objective of this study was to compare the performance of base-level and meta-level classifiers on the task of physical activity recognition. Five wireless kinematic sensors were attached to each subject (n = 25) while they completed a range of basic physical activities in a controlled laboratory setting. Subjects were then asked to carry out similar self-annotated physical activities in a random order and in an unsupervised environment. A combination of time-domain and frequency-domain features were extracted from the sensor data including the first four central moments, zero-crossing rate, average magnitude, sensor cross-correlation, sensor auto-correlation, spectral entropy and dominant frequency components. A reduced feature set was generated using a wrapper subset evaluation technique with a linear forward search and this feature set was employed for classifier comparison. The meta-level classifier AdaBoostM1 with C4.5 Graft as its base-level classifier achieved an overall accuracy of 95%. Equal sized datasets of subject independent data and subject dependent data were used to train this classifier and high recognition rates could be achieved without the need for user specific training. Furthermore, it was found that an accuracy of 88% could be achieved using data from the ankle and wrist sensors only.

Collected Notes on the Workshop for Pattern Discovery in Large Databases

NASA Technical Reports Server (NTRS)

Buntine, Wray (Editor); Delalto, Martha (Editor)

1991-01-01

These collected notes are a record of material presented at the Workshop. The core data analysis is addressed that have traditionally required statistical or pattern recognition techniques. Some of the core tasks include classification, discrimination, clustering, supervised and unsupervised learning, discovery and diagnosis, i.e., general pattern discovery.

Unsupervised method for automatic construction of a disease dictionary from a large free text collection.

PubMed

Xu, Rong; Supekar, Kaustubh; Morgan, Alex; Das, Amar; Garber, Alan

2008-11-06

Concept specific lexicons (e.g. diseases, drugs, anatomy) are a critical source of background knowledge for many medical language-processing systems. However, the rapid pace of biomedical research and the lack of constraints on usage ensure that such dictionaries are incomplete. Focusing on disease terminology, we have developed an automated, unsupervised, iterative pattern learning approach for constructing a comprehensive medical dictionary of disease terms from randomized clinical trial (RCT) abstracts, and we compared different ranking methods for automatically extracting con-textual patterns and concept terms. When used to identify disease concepts from 100 randomly chosen, manually annotated clinical abstracts, our disease dictionary shows significant performance improvement (F1 increased by 35-88%) over available, manually created disease terminologies.

Unsupervised Method for Automatic Construction of a Disease Dictionary from a Large Free Text Collection

PubMed Central

Xu, Rong; Supekar, Kaustubh; Morgan, Alex; Das, Amar; Garber, Alan

2008-01-01

Concept specific lexicons (e.g. diseases, drugs, anatomy) are a critical source of background knowledge for many medical language-processing systems. However, the rapid pace of biomedical research and the lack of constraints on usage ensure that such dictionaries are incomplete. Focusing on disease terminology, we have developed an automated, unsupervised, iterative pattern learning approach for constructing a comprehensive medical dictionary of disease terms from randomized clinical trial (RCT) abstracts, and we compared different ranking methods for automatically extracting contextual patterns and concept terms. When used to identify disease concepts from 100 randomly chosen, manually annotated clinical abstracts, our disease dictionary shows significant performance improvement (F1 increased by 35–88%) over available, manually created disease terminologies. PMID:18999169

Learning and Tuning of Fuzzy Rules

NASA Technical Reports Server (NTRS)

Berenji, Hamid R.

1997-01-01

In this chapter, we review some of the current techniques for learning and tuning fuzzy rules. For clarity, we refer to the process of generating rules from data as the learning problem and distinguish it from tuning an already existing set of fuzzy rules. For learning, we touch on unsupervised learning techniques such as fuzzy c-means, fuzzy decision tree systems, fuzzy genetic algorithms, and linear fuzzy rules generation methods. For tuning, we discuss Jang's ANFIS architecture, Berenji-Khedkar's GARIC architecture and its extensions in GARIC-Q. We show that the hybrid techniques capable of learning and tuning fuzzy rules, such as CART-ANFIS, RNN-FLCS, and GARIC-RB, are desirable in development of a number of future intelligent systems.

Cascaded ensemble of convolutional neural networks and handcrafted features for mitosis detection

NASA Astrophysics Data System (ADS)

Wang, Haibo; Cruz-Roa, Angel; Basavanhally, Ajay; Gilmore, Hannah; Shih, Natalie; Feldman, Mike; Tomaszewski, John; Gonzalez, Fabio; Madabhushi, Anant

2014-03-01

Breast cancer (BCa) grading plays an important role in predicting disease aggressiveness and patient outcome. A key component of BCa grade is mitotic count, which involves quantifying the number of cells in the process of dividing (i.e. undergoing mitosis) at a specific point in time. Currently mitosis counting is done manually by a pathologist looking at multiple high power fields on a glass slide under a microscope, an extremely laborious and time consuming process. The development of computerized systems for automated detection of mitotic nuclei, while highly desirable, is confounded by the highly variable shape and appearance of mitoses. Existing methods use either handcrafted features that capture certain morphological, statistical or textural attributes of mitoses or features learned with convolutional neural networks (CNN). While handcrafted features are inspired by the domain and the particular application, the data-driven CNN models tend to be domain agnostic and attempt to learn additional feature bases that cannot be represented through any of the handcrafted features. On the other hand, CNN is computationally more complex and needs a large number of labeled training instances. Since handcrafted features attempt to model domain pertinent attributes and CNN approaches are largely unsupervised feature generation methods, there is an appeal to attempting to combine these two distinct classes of feature generation strategies to create an integrated set of attributes that can potentially outperform either class of feature extraction strategies individually. In this paper, we present a cascaded approach for mitosis detection that intelligently combines a CNN model and handcrafted features (morphology, color and texture features). By employing a light CNN model, the proposed approach is far less demanding computationally, and the cascaded strategy of combining handcrafted features and CNN-derived features enables the possibility of maximizing performance by leveraging the disconnected feature sets. Evaluation on the public ICPR12 mitosis dataset that has 226 mitoses annotated on 35 High Power Fields (HPF, x400 magnification) by several pathologists and 15 testing HPFs yielded an F-measure of 0.7345. Apart from this being the second best performance ever recorded for this MITOS dataset, our approach is faster and requires fewer computing resources compared to extant methods, making this feasible for clinical use.

Neural signal processing and closed-loop control algorithm design for an implanted neural recording and stimulation system.

PubMed

Hamilton, Lei; McConley, Marc; Angermueller, Kai; Goldberg, David; Corba, Massimiliano; Kim, Louis; Moran, James; Parks, Philip D; Sang Chin; Widge, Alik S; Dougherty, Darin D; Eskandar, Emad N

2015-08-01

A fully autonomous intracranial device is built to continually record neural activities in different parts of the brain, process these sampled signals, decode features that correlate to behaviors and neuropsychiatric states, and use these features to deliver brain stimulation in a closed-loop fashion. In this paper, we describe the sampling and stimulation aspects of such a device. We first describe the signal processing algorithms of two unsupervised spike sorting methods. Next, we describe the LFP time-frequency analysis and feature derivation from the two spike sorting methods. Spike sorting includes a novel approach to constructing a dictionary learning algorithm in a Compressed Sensing (CS) framework. We present a joint prediction scheme to determine the class of neural spikes in the dictionary learning framework; and, the second approach is a modified OSort algorithm which is implemented in a distributed system optimized for power efficiency. Furthermore, sorted spikes and time-frequency analysis of LFP signals can be used to generate derived features (including cross-frequency coupling, spike-field coupling). We then show how these derived features can be used in the design and development of novel decode and closed-loop control algorithms that are optimized to apply deep brain stimulation based on a patient's neuropsychiatric state. For the control algorithm, we define the state vector as representative of a patient's impulsivity, avoidance, inhibition, etc. Controller parameters are optimized to apply stimulation based on the state vector's current state as well as its historical values. The overall algorithm and software design for our implantable neural recording and stimulation system uses an innovative, adaptable, and reprogrammable architecture that enables advancement of the state-of-the-art in closed-loop neural control while also meeting the challenges of system power constraints and concurrent development with ongoing scientific research designed to define brain network connectivity and neural network dynamics that vary at the individual patient level and vary over time.

Adaptive Neuron Apoptosis for Accelerating Deep Learning on Large Scale Systems

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

Siegel, Charles M.; Daily, Jeffrey A.; Vishnu, Abhinav

Machine Learning and Data Mining (MLDM) algorithms are becoming ubiquitous in {\\em model learning} from the large volume of data generated using simulations, experiments and handheld devices. Deep Learning algorithms -- a class of MLDM algorithms -- are applied for automatic feature extraction, and learning non-linear models for unsupervised and supervised algorithms. Naturally, several libraries which support large scale Deep Learning -- such as TensorFlow and Caffe -- have become popular. In this paper, we present novel techniques to accelerate the convergence of Deep Learning algorithms by conducting low overhead removal of redundant neurons -- {\\em apoptosis} of neurons --more » which do not contribute to model learning, during the training phase itself. We provide in-depth theoretical underpinnings of our heuristics (bounding accuracy loss and handling apoptosis of several neuron types), and present the methods to conduct adaptive neuron apoptosis. We implement our proposed heuristics with the recently introduced TensorFlow and using its recently proposed extension with MPI. Our performance evaluation on two difference clusters -- one connected with Intel Haswell multi-core systems, and other with nVIDIA GPUs -- using InfiniBand, indicates the efficacy of the proposed heuristics and implementations. Specifically, we are able to improve the training time for several datasets by 2-3x, while reducing the number of parameters by 30x (4-5x on average) on datasets such as ImageNet classification. For the Higgs Boson dataset, our implementation improves the accuracy (measured by Area Under Curve (AUC)) for classification from 0.88/1 to 0.94/1, while reducing the number of parameters by 3x in comparison to existing literature, while achieving a 2.44x speedup in comparison to the default (no apoptosis) algorithm.« less

Conditional High-Order Boltzmann Machines for Supervised Relation Learning.

PubMed

Huang, Yan; Wang, Wei; Wang, Liang; Tan, Tieniu

2017-09-01

Relation learning is a fundamental problem in many vision tasks. Recently, high-order Boltzmann machine and its variants have shown their great potentials in learning various types of data relation in a range of tasks. But most of these models are learned in an unsupervised way, i.e., without using relation class labels, which are not very discriminative for some challenging tasks, e.g., face verification. In this paper, with the goal to perform supervised relation learning, we introduce relation class labels into conventional high-order multiplicative interactions with pairwise input samples, and propose a conditional high-order Boltzmann Machine (CHBM), which can learn to classify the data relation in a binary classification way. To be able to deal with more complex data relation, we develop two improved variants of CHBM: 1) latent CHBM, which jointly performs relation feature learning and classification, by using a set of latent variables to block the pathway from pairwise input samples to output relation labels and 2) gated CHBM, which untangles factors of variation in data relation, by exploiting a set of latent variables to multiplicatively gate the classification of CHBM. To reduce the large number of model parameters generated by the multiplicative interactions, we approximately factorize high-order parameter tensors into multiple matrices. Then, we develop efficient supervised learning algorithms, by first pretraining the models using joint likelihood to provide good parameter initialization, and then finetuning them using conditional likelihood to enhance the discriminant ability. We apply the proposed models to a series of tasks including invariant recognition, face verification, and action similarity labeling. Experimental results demonstrate that by exploiting supervised relation labels, our models can greatly improve the performance.

Unsupervised learning in general connectionist systems.

PubMed

Dente, J A; Mendes, R Vilela

1996-01-01

There is a common framework in which different connectionist systems may be treated in a unified way. The general system in which they may all be mapped is a network which, in addition to the connection strengths, has an adaptive node parameter controlling the output intensity. In this paper we generalize two neural network learning schemes to networks with node parameters. In generalized Hebbian learning we find improvements to the convergence rate for small eigenvalues in principal component analysis. For competitive learning the use of node parameters also seems useful in that, by emphasizing or de-emphasizing the dominance of winning neurons, either improved robustness or discrimination is obtained.

Machine learning applications in genetics and genomics.

PubMed

Libbrecht, Maxwell W; Noble, William Stafford

2015-06-01

The field of machine learning, which aims to develop computer algorithms that improve with experience, holds promise to enable computers to assist humans in the analysis of large, complex data sets. Here, we provide an overview of machine learning applications for the analysis of genome sequencing data sets, including the annotation of sequence elements and epigenetic, proteomic or metabolomic data. We present considerations and recurrent challenges in the application of supervised, semi-supervised and unsupervised machine learning methods, as well as of generative and discriminative modelling approaches. We provide general guidelines to assist in the selection of these machine learning methods and their practical application for the analysis of genetic and genomic data sets.

A Case Study on Sepsis Using PubMed and Deep Learning for Ontology Learning.

PubMed

Arguello Casteleiro, Mercedes; Maseda Fernandez, Diego; Demetriou, George; Read, Warren; Fernandez Prieto, Maria Jesus; Des Diz, Julio; Nenadic, Goran; Keane, John; Stevens, Robert

2017-01-01

We investigate the application of distributional semantics models for facilitating unsupervised extraction of biomedical terms from unannotated corpora. Term extraction is used as the first step of an ontology learning process that aims to (semi-)automatic annotation of biomedical concepts and relations from more than 300K PubMed titles and abstracts. We experimented with both traditional distributional semantics methods such as Latent Semantic Analysis (LSA) and Latent Dirichlet Allocation (LDA) as well as the neural language models CBOW and Skip-gram from Deep Learning. The evaluation conducted concentrates on sepsis, a major life-threatening condition, and shows that Deep Learning models outperform LSA and LDA with much higher precision.

LIA at TREC 2012 Web Track: Unsupervised Search Concepts Identification from General Sources of Information

DTIC Science & Technology

2012-11-01

use in this work the variational approximation algo- rithm implemented and distributed by Pr . Blei1. Each learned multinomial distribution φk is tra...4,111,240 newswire articles collected from four distinct international sources including the New York Times (Graff and Cieri, 2003). The New York Times

Promising Ideas for Collective Advancement of Communal Knowledge Using Temporal Analytics and Cluster Analysis

ERIC Educational Resources Information Center

Lee, Alwyn Vwen Yen; Tan, Seng Chee

2017-01-01

Understanding ideas in a discourse is challenging, especially in textual discourse analysis. We propose using temporal analytics with unsupervised machine learning techniques to investigate promising ideas for the collective advancement of communal knowledge in an online knowledge building discourse. A discourse unit network was constructed and…

An Empirical Generative Framework for Computational Modeling of Language Acquisition

ERIC Educational Resources Information Center

Waterfall, Heidi R.; Sandbank, Ben; Onnis, Luca; Edelman, Shimon

2010-01-01

This paper reports progress in developing a computer model of language acquisition in the form of (1) a generative grammar that is (2) algorithmically learnable from realistic corpus data, (3) viable in its large-scale quantitative performance and (4) psychologically real. First, we describe new algorithmic methods for unsupervised learning of…

High-Dimensional Semantic Space Accounts of Priming

ERIC Educational Resources Information Center

Jones, Michael N.; Kintsch, Walter; Mewhort, Douglas J. K.

2006-01-01

A broad range of priming data has been used to explore the structure of semantic memory and to test between models of word representation. In this paper, we examine the computational mechanisms required to learn distributed semantic representations for words directly from unsupervised experience with language. To best account for the variety of…

Alternative to Proctoring in Introductory Statistics Community College Courses

ERIC Educational Resources Information Center

Feinman, Yalena

2018-01-01

The credibility of unsupervised exams, one of the biggest challenges of e-learning, is currently maintained by proctoring. However, little has been done to determine whether expensive and inconvenient proctoring is necessary. The purpose of this quantitative study was to determine whether the use of security mechanisms, based on the taxonomy of…

Robust Arm and Hand Tracking by Unsupervised Context Learning

PubMed Central

Spruyt, Vincent; Ledda, Alessandro; Philips, Wilfried

2014-01-01

Hand tracking in video is an increasingly popular research field due to the rise of novel human-computer interaction methods. However, robust and real-time hand tracking in unconstrained environments remains a challenging task due to the high number of degrees of freedom and the non-rigid character of the human hand. In this paper, we propose an unsupervised method to automatically learn the context in which a hand is embedded. This context includes the arm and any other object that coherently moves along with the hand. We introduce two novel methods to incorporate this context information into a probabilistic tracking framework, and introduce a simple yet effective solution to estimate the position of the arm. Finally, we show that our method greatly increases robustness against occlusion and cluttered background, without degrading tracking performance if no contextual information is available. The proposed real-time algorithm is shown to outperform the current state-of-the-art by evaluating it on three publicly available video datasets. Furthermore, a novel dataset is created and made publicly available for the research community. PMID:25004155

Quick fuzzy backpropagation algorithm.

PubMed

Nikov, A; Stoeva, S

2001-03-01

A modification of the fuzzy backpropagation (FBP) algorithm called QuickFBP algorithm is proposed, where the computation of the net function is significantly quicker. It is proved that the FBP algorithm is of exponential time complexity, while the QuickFBP algorithm is of polynomial time complexity. Convergence conditions of the QuickFBP, resp. the FBP algorithm are defined and proved for: (1) single output neural networks in case of training patterns with different targets; and (2) multiple output neural networks in case of training patterns with equivalued target vector. They support the automation of the weights training process (quasi-unsupervised learning) establishing the target value(s) depending on the network's input values. In these cases the simulation results confirm the convergence of both algorithms. An example with a large-sized neural network illustrates the significantly greater training speed of the QuickFBP rather than the FBP algorithm. The adaptation of an interactive web system to users on the basis of the QuickFBP algorithm is presented. Since the QuickFBP algorithm ensures quasi-unsupervised learning, this implies its broad applicability in areas of adaptive and adaptable interactive systems, data mining, etc. applications.

Deep neural ensemble for retinal vessel segmentation in fundus images towards achieving label-free angiography.

PubMed

Lahiri, A; Roy, Abhijit Guha; Sheet, Debdoot; Biswas, Prabir Kumar

2016-08-01

Automated segmentation of retinal blood vessels in label-free fundus images entails a pivotal role in computed aided diagnosis of ophthalmic pathologies, viz., diabetic retinopathy, hypertensive disorders and cardiovascular diseases. The challenge remains active in medical image analysis research due to varied distribution of blood vessels, which manifest variations in their dimensions of physical appearance against a noisy background. In this paper we formulate the segmentation challenge as a classification task. Specifically, we employ unsupervised hierarchical feature learning using ensemble of two level of sparsely trained denoised stacked autoencoder. First level training with bootstrap samples ensures decoupling and second level ensemble formed by different network architectures ensures architectural revision. We show that ensemble training of auto-encoders fosters diversity in learning dictionary of visual kernels for vessel segmentation. SoftMax classifier is used for fine tuning each member autoencoder and multiple strategies are explored for 2-level fusion of ensemble members. On DRIVE dataset, we achieve maximum average accuracy of 95.33% with an impressively low standard deviation of 0.003 and Kappa agreement coefficient of 0.708. Comparison with other major algorithms substantiates the high efficacy of our model.

Quantification of Hepatorenal Index for Computer-Aided Fatty Liver Classification with Self-Organizing Map and Fuzzy Stretching from Ultrasonography.

PubMed

Kim, Kwang Baek; Kim, Chang Won

2015-01-01

Accurate measures of liver fat content are essential for investigating hepatic steatosis. For a noninvasive inexpensive ultrasonographic analysis, it is necessary to validate the quantitative assessment of liver fat content so that fully automated reliable computer-aided software can assist medical practitioners without any operator subjectivity. In this study, we attempt to quantify the hepatorenal index difference between the liver and the kidney with respect to the multiple severity status of hepatic steatosis. In order to do this, a series of carefully designed image processing techniques, including fuzzy stretching and edge tracking, are applied to extract regions of interest. Then, an unsupervised neural learning algorithm, the self-organizing map, is designed to establish characteristic clusters from the image, and the distribution of the hepatorenal index values with respect to the different levels of the fatty liver status is experimentally verified to estimate the differences in the distribution of the hepatorenal index. Such findings will be useful in building reliable computer-aided diagnostic software if combined with a good set of other characteristic feature sets and powerful machine learning classifiers in the future.

Quantification of Hepatorenal Index for Computer-Aided Fatty Liver Classification with Self-Organizing Map and Fuzzy Stretching from Ultrasonography

PubMed Central

Kim, Kwang Baek

2015-01-01

Accurate measures of liver fat content are essential for investigating hepatic steatosis. For a noninvasive inexpensive ultrasonographic analysis, it is necessary to validate the quantitative assessment of liver fat content so that fully automated reliable computer-aided software can assist medical practitioners without any operator subjectivity. In this study, we attempt to quantify the hepatorenal index difference between the liver and the kidney with respect to the multiple severity status of hepatic steatosis. In order to do this, a series of carefully designed image processing techniques, including fuzzy stretching and edge tracking, are applied to extract regions of interest. Then, an unsupervised neural learning algorithm, the self-organizing map, is designed to establish characteristic clusters from the image, and the distribution of the hepatorenal index values with respect to the different levels of the fatty liver status is experimentally verified to estimate the differences in the distribution of the hepatorenal index. Such findings will be useful in building reliable computer-aided diagnostic software if combined with a good set of other characteristic feature sets and powerful machine learning classifiers in the future. PMID:26247023

Exploiting range imagery: techniques and applications

NASA Astrophysics Data System (ADS)

Armbruster, Walter

2009-07-01

Practically no applications exist for which automatic processing of 2D intensity imagery can equal human visual perception. This is not the case for range imagery. The paper gives examples of 3D laser radar applications, for which automatic data processing can exceed human visual cognition capabilities and describes basic processing techniques for attaining these results. The examples are drawn from the fields of helicopter obstacle avoidance, object detection in surveillance applications, object recognition at high range, multi-object-tracking, and object re-identification in range image sequences. Processing times and recognition performances are summarized. The techniques used exploit the bijective continuity of the imaging process as well as its independence of object reflectivity, emissivity and illumination. This allows precise formulations of the probability distributions involved in figure-ground segmentation, feature-based object classification and model based object recognition. The probabilistic approach guarantees optimal solutions for single images and enables Bayesian learning in range image sequences. Finally, due to recent results in 3D-surface completion, no prior model libraries are required for recognizing and re-identifying objects of quite general object categories, opening the way to unsupervised learning and fully autonomous cognitive systems.

A compound memristive synapse model for statistical learning through STDP in spiking neural networks

PubMed Central

Bill, Johannes; Legenstein, Robert

2014-01-01

Memristors have recently emerged as promising circuit elements to mimic the function of biological synapses in neuromorphic computing. The fabrication of reliable nanoscale memristive synapses, that feature continuous conductance changes based on the timing of pre- and postsynaptic spikes, has however turned out to be challenging. In this article, we propose an alternative approach, the compound memristive synapse, that circumvents this problem by the use of memristors with binary memristive states. A compound memristive synapse employs multiple bistable memristors in parallel to jointly form one synapse, thereby providing a spectrum of synaptic efficacies. We investigate the computational implications of synaptic plasticity in the compound synapse by integrating the recently observed phenomenon of stochastic filament formation into an abstract model of stochastic switching. Using this abstract model, we first show how standard pulsing schemes give rise to spike-timing dependent plasticity (STDP) with a stabilizing weight dependence in compound synapses. In a next step, we study unsupervised learning with compound synapses in networks of spiking neurons organized in a winner-take-all architecture. Our theoretical analysis reveals that compound-synapse STDP implements generalized Expectation-Maximization in the spiking network. Specifically, the emergent synapse configuration represents the most salient features of the input distribution in a Mixture-of-Gaussians generative model. Furthermore, the network's spike response to spiking input streams approximates a well-defined Bayesian posterior distribution. We show in computer simulations how such networks learn to represent high-dimensional distributions over images of handwritten digits with high fidelity even in presence of substantial device variations and under severe noise conditions. Therefore, the compound memristive synapse may provide a synaptic design principle for future neuromorphic architectures. PMID:25565943

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

Parekh, V; Jacobs, MA

Purpose: Multiparametric radiological imaging is used for diagnosis in patients. Potentially extracting useful features specific to a patient’s pathology would be crucial step towards personalized medicine and assessing treatment options. In order to automatically extract features directly from multiparametric radiological imaging datasets, we developed an advanced unsupervised machine learning algorithm called the multidimensional imaging radiomics-geodesics(MIRaGe). Methods: Seventy-six breast tumor patients underwent 3T MRI breast imaging were used for this study. We tested the MIRaGe algorithm to extract features for classification of breast tumors into benign or malignant. The MRI parameters used were T1-weighted, T2-weighted, dynamic contrast enhanced MR imaging (DCE-MRI)more » and diffusion weighted imaging(DWI). The MIRaGe algorithm extracted the radiomics-geodesics features (RGFs) from multiparametric MRI datasets. This enable our method to learn the intrinsic manifold representations corresponding to the patients. To determine the informative RGF, a modified Isomap algorithm(t-Isomap) was created for a radiomics-geodesics feature space(tRGFS) to avoid overfitting. Final classification was performed using SVM. The predictive power of the RGFs was tested and validated using k-fold cross validation. Results: The RGFs extracted by the MIRaGe algorithm successfully classified malignant lesions from benign lesions with a sensitivity of 93% and a specificity of 91%. The top 50 RGFs identified as the most predictive by the t-Isomap procedure were consistent with the radiological parameters known to be associated with breast cancer diagnosis and were categorized as kinetic curve characterizing RGFs, wash-in rate characterizing RGFs, wash-out rate characterizing RGFs and morphology characterizing RGFs. Conclusion: In this paper, we developed a novel feature extraction algorithm for multiparametric radiological imaging. The results demonstrated the power of the MIRaGe algorithm at automatically discovering useful feature representations directly from the raw multiparametric MRI data. In conclusion, the MIRaGe informatics model provides a powerful tool with applicability in cancer diagnosis and a possibility of extension to other kinds of pathologies. NIH (P50CA103175, 5P30CA006973 (IRAT), R01CA190299, U01CA140204), Siemens Medical Systems (JHU-2012-MR-86-01) and Nivida Graphics Corporation.« less

Rectified factor networks for biclustering of omics data.

PubMed

Clevert, Djork-Arné; Unterthiner, Thomas; Povysil, Gundula; Hochreiter, Sepp

2017-07-15

Biclustering has become a major tool for analyzing large datasets given as matrix of samples times features and has been successfully applied in life sciences and e-commerce for drug design and recommender systems, respectively. actor nalysis for cluster cquisition (FABIA), one of the most successful biclustering methods, is a generative model that represents each bicluster by two sparse membership vectors: one for the samples and one for the features. However, FABIA is restricted to about 20 code units because of the high computational complexity of computing the posterior. Furthermore, code units are sometimes insufficiently decorrelated and sample membership is difficult to determine. We propose to use the recently introduced unsupervised Deep Learning approach Rectified Factor Networks (RFNs) to overcome the drawbacks of existing biclustering methods. RFNs efficiently construct very sparse, non-linear, high-dimensional representations of the input via their posterior means. RFN learning is a generalized alternating minimization algorithm based on the posterior regularization method which enforces non-negative and normalized posterior means. Each code unit represents a bicluster, where samples for which the code unit is active belong to the bicluster and features that have activating weights to the code unit belong to the bicluster. On 400 benchmark datasets and on three gene expression datasets with known clusters, RFN outperformed 13 other biclustering methods including FABIA. On data of the 1000 Genomes Project, RFN could identify DNA segments which indicate, that interbreeding with other hominins starting already before ancestors of modern humans left Africa. https://github.com/bioinf-jku/librfn. djork-arne.clevert@bayer.com or hochreit@bioinf.jku.at. © The Author 2017. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Semi-automatic ground truth generation using unsupervised clustering and limited manual labeling: Application to handwritten character recognition

PubMed Central

Vajda, Szilárd; Rangoni, Yves; Cecotti, Hubert

2015-01-01

For training supervised classifiers to recognize different patterns, large data collections with accurate labels are necessary. In this paper, we propose a generic, semi-automatic labeling technique for large handwritten character collections. In order to speed up the creation of a large scale ground truth, the method combines unsupervised clustering and minimal expert knowledge. To exploit the potential discriminant complementarities across features, each character is projected into five different feature spaces. After clustering the images in each feature space, the human expert labels the cluster centers. Each data point inherits the label of its cluster’s center. A majority (or unanimity) vote decides the label of each character image. The amount of human involvement (labeling) is strictly controlled by the number of clusters – produced by the chosen clustering approach. To test the efficiency of the proposed approach, we have compared, and evaluated three state-of-the art clustering methods (k-means, self-organizing maps, and growing neural gas) on the MNIST digit data set, and a Lampung Indonesian character data set, respectively. Considering a k-nn classifier, we show that labeling manually only 1.3% (MNIST), and 3.2% (Lampung) of the training data, provides the same range of performance than a completely labeled data set would. PMID:25870463

ISO learning approximates a solution to the inverse-controller problem in an unsupervised behavioral paradigm.

PubMed

Porr, Bernd; von Ferber, Christian; Wörgötter, Florentin

2003-04-01

In "Isotropic Sequence Order Learning" (pp. 831-864 in this issue), we introduced a novel algorithm for temporal sequence learning (ISO learning). Here, we embed this algorithm into a formal nonevaluating (teacher free) environment, which establishes a sensor-motor feedback. The system is initially guided by a fixed reflex reaction, which has the objective disadvantage that it can react only after a disturbance has occurred. ISO learning eliminates this disadvantage by replacing the reflex-loop reactions with earlier anticipatory actions. In this article, we analytically demonstrate that this process can be understood in terms of control theory, showing that the system learns the inverse controller of its own reflex. Thereby, this system is able to learn a simple form of feedforward motor control.

LacSubPred: predicting subtypes of Laccases, an important lignin metabolism-related enzyme class, using in silico approaches

PubMed Central

2014-01-01

Background Laccases (E.C. 1.10.3.2) are multi-copper oxidases that have gained importance in many industries such as biofuels, pulp production, textile dye bleaching, bioremediation, and food production. Their usefulness stems from the ability to act on a diverse range of phenolic compounds such as o-/p-quinols, aminophenols, polyphenols, polyamines, aryl diamines, and aromatic thiols. Despite acting on a wide range of compounds as a family, individual Laccases often exhibit distinctive and varied substrate ranges. This is likely due to Laccases involvement in many metabolic roles across diverse taxa. Classification systems for multi-copper oxidases have been developed using multiple sequence alignments, however, these systems seem to largely follow species taxonomy rather than substrate ranges, enzyme properties, or specific function. It has been suggested that the roles and substrates of various Laccases are related to their optimal pH. This is consistent with the observation that fungal Laccases usually prefer acidic conditions, whereas plant and bacterial Laccases prefer basic conditions. Based on these observations, we hypothesize that a descriptor-based unsupervised learning system could generate homology independent classification system for better describing the functional properties of Laccases. Results In this study, we first utilized unsupervised learning approach to develop a novel homology independent Laccase classification system. From the descriptors considered, physicochemical properties showed the best performance. Physicochemical properties divided the Laccases into twelve subtypes. Analysis of the clusters using a t-test revealed that the majority of the physicochemical descriptors had statistically significant differences between the classes. Feature selection identified the most important features as negatively charges residues, the peptide isoelectric point, and acidic or amidic residues. Secondly, to allow for classification of new Laccases, a supervised learning system was developed from the clusters. The models showed high performance with an overall accuracy of 99.03%, error of 0.49%, MCC of 0.9367, precision of 94.20%, sensitivity of 94.20%, and specificity of 99.47% in a 5-fold cross-validation test. In an independent test, our models still provide a high accuracy of 97.98%, error rate of 1.02%, MCC of 0.8678, precision of 87.88%, sensitivity of 87.88% and specificity of 98.90%. Conclusion This study provides a useful classification system for better understanding of Laccases from their physicochemical properties perspective. We also developed a publically available web tool for the characterization of Laccase protein sequences (http://lacsubpred.bioinfo.ucr.edu/). Finally, the programs used in the study are made available for researchers interested in applying the system to other enzyme classes (https://github.com/tweirick/SubClPred). PMID:25350584

VHR satellite multitemporal data to extract cultural landscape changes in the roman site of Grumentum

NASA Astrophysics Data System (ADS)

masini, nicola; Lasaponara, Rosa

2013-04-01

The papers deals with the use of VHR satellite multitemporal data set to extract cultural landscape changes in the roman site of Grumentum Grumentum is an ancient town, 50 km south of Potenza, located near the roman road of Via Herculea which connected the Venusia, in the north est of Basilicata, with Heraclea in the Ionian coast. The first settlement date back to the 6th century BC. It was resettled by the Romans in the 3rd century BC. Its urban fabric which evidences a long history from the Republican age to late Antiquity (III BC-V AD) is composed of the typical urban pattern of cardi and decumani. Its excavated ruins include a large amphitheatre, a theatre, the thermae, the Forum and some temples. There are many techniques nowadays available to capture and record differences in two or more images. In this paper we focus and apply the two main approaches which can be distinguished into : (i) unsupervised and (ii) supervised change detection methods. Unsupervised change detection methods are generally based on the transformation of the two multispectral images in to a single band or multiband image which are further analyzed to identify changes Unsupervised change detection techniques are generally based on three basic steps (i) the preprocessing step, (ii) a pixel-by-pixel comparison is performed, (iii). Identification of changes according to the magnitude an direction (positive /negative). Unsupervised change detection are generally based on the transformation of the two multispectral images into a single band or multiband image which are further analyzed to identify changes. Than the separation between changed and unchanged classes is obtained from the magnitude of the resulting spectral change vectors by means of empirical or theoretical well founded approaches Supervised change detection methods are generally based on supervised classification methods, which require the availability of a suitable training set for the learning process of the classifiers. Unsupervised change detection techniques are generally based on three basic steps (i) the preprocessing step, (ii) supervised classification is performed on the single dates or on the map obtained as the difference of two dates, (iii). Identification of changes according to the magnitude an direction (positive /negative). Supervised change detection are generally based on supervised classification methods, which require the availability of a suitable training set for the learning process of the classifiers, therefore these algorithms require a preliminary knowledge necessary: (i) to generate representative parameters for each class of interest; and (ii) to carry out the training stage Advantages and disadvantages of the supervised and unsupervised approaches are discuss. Finally results from the the satellite multitemporal dataset was also integrated with aerial photos from historical archive in order to expand the time window of the investigation and capture landscape changes occurred from the Agrarian Reform, in the 50s, up today.

Subtyping of Children with Developmental Dyslexia via Bootstrap Aggregated Clustering and the Gap Statistic: Comparison with the Double-Deficit Hypothesis

ERIC Educational Resources Information Center

King, Wayne M.; Giess, Sally A.; Lombardino, Linda J.

2007-01-01

Background: The marked degree of heterogeneity in persons with developmental dyslexia has motivated the investigation of possible subtypes. Attempts have proceeded both from theoretical models of reading and the application of unsupervised learning (clustering) methods. Previous cluster analyses of data obtained from persons with reading…

Featureless classification of light curves

NASA Astrophysics Data System (ADS)

Kügler, S. D.; Gianniotis, N.; Polsterer, K. L.

2015-08-01

In the era of rapidly increasing amounts of time series data, classification of variable objects has become the main objective of time-domain astronomy. Classification of irregularly sampled time series is particularly difficult because the data cannot be represented naturally as a vector which can be directly fed into a classifier. In the literature, various statistical features serve as vector representations. In this work, we represent time series by a density model. The density model captures all the information available, including measurement errors. Hence, we view this model as a generalization to the static features which directly can be derived, e.g. as moments from the density. Similarity between each pair of time series is quantified by the distance between their respective models. Classification is performed on the obtained distance matrix. In the numerical experiments, we use data from the OGLE (Optical Gravitational Lensing Experiment) and ASAS (All Sky Automated Survey) surveys and demonstrate that the proposed representation performs up to par with the best currently used feature-based approaches. The density representation preserves all static information present in the observational data, in contrast to a less-complete description by features. The density representation is an upper boundary in terms of information made available to the classifier. Consequently, the predictive power of the proposed classification depends on the choice of similarity measure and classifier, only. Due to its principled nature, we advocate that this new approach of representing time series has potential in tasks beyond classification, e.g. unsupervised learning.

Class imbalance in unsupervised change detection - A diagnostic analysis from urban remote sensing

NASA Astrophysics Data System (ADS)

Leichtle, Tobias; Geiß, Christian; Lakes, Tobia; Taubenböck, Hannes

2017-08-01

Automatic monitoring of changes on the Earth's surface is an intrinsic capability and simultaneously a persistent methodological challenge in remote sensing, especially regarding imagery with very-high spatial resolution (VHR) and complex urban environments. In order to enable a high level of automatization, the change detection problem is solved in an unsupervised way to alleviate efforts associated with collection of properly encoded prior knowledge. In this context, this paper systematically investigates the nature and effects of class distribution and class imbalance in an unsupervised binary change detection application based on VHR imagery over urban areas. For this purpose, a diagnostic framework for sensitivity analysis of a large range of possible degrees of class imbalance is presented, which is of particular importance with respect to unsupervised approaches where the content of images and thus the occurrence and the distribution of classes are generally unknown a priori. Furthermore, this framework can serve as a general technique to evaluate model transferability in any two-class classification problem. The applied change detection approach is based on object-based difference features calculated from VHR imagery and subsequent unsupervised two-class clustering using k-means, genetic k-means and self-organizing map (SOM) clustering. The results from two test sites with different structural characteristics of the built environment demonstrated that classification performance is generally worse in imbalanced class distribution settings while best results were reached in balanced or close to balanced situations. Regarding suitable accuracy measures for evaluating model performance in imbalanced settings, this study revealed that the Kappa statistics show significant response to class distribution while the true skill statistic was widely insensitive to imbalanced classes. In general, the genetic k-means clustering algorithm achieved the most robust results with respect to class imbalance while the SOM clustering exhibited a distinct optimization towards a balanced distribution of classes.