The 2013 International Joint Conference on Neural Networks (IJCNN), 2013
ABSTRACT Increasing number of practical applications that involve streaming nonstationary data ha... more ABSTRACT Increasing number of practical applications that involve streaming nonstationary data have led to a recent surge in algorithms designed to learn from such data. One challenging version of this problem that has not received as much attention, however, is learning streaming nonstationary data when a small initial set of data are labeled, with unlabeled data being available thereafter. We have recently introduced the COMPOSE algorithm for learning in such scenarios, which we refer to as initially labeled nonstationary streaming data. COMPOSE works remarkably well, however it requires limited (gradual) drift, and cannot address special cases such as introduction of a new class or significant overlap of existing classes, as such scenarios cannot be learned without additional labeled data. Scenarios that provide occasional or periodic limited labeled data are not uncommon, however, for which many of COMPOSE's restrictions can be lifted. In this contribution, we describe a new version of COMPOSE as a proof-of-concept algorithm that can identify the instances whose labels — if available — would be most beneficial, and then combine those instances with unlabeled data to actively learn from streaming nonstationary data, even when the distribution of the data experiences abrupt changes. On two carefully designed experiments that include abrupt changes as well as addition of new classes, we show that COMPOSE.AL significantly outperforms original COMPOSE, while closely tracking the optimal Bayes classifier performance.
— Selection of most informative features that leads to a small loss on future data are arguably o... more — Selection of most informative features that leads to a small loss on future data are arguably one of the most important steps in classification, data analysis and model selection. Several feature selection (FS) algorithms are available; however, due to noise present in any data set, FS algorithms are typically accompanied by an appropriate cross-validation scheme. In this brief, we propose a statistical hypothesis test derived from the Neyman–Pearson lemma for determining if a feature is statistically relevant. The proposed approach can be applied as a wrapper to any FS algorithm, regardless of the FS criteria used by that algorithm, to determine whether a feature belongs in the relevant set. Perhaps more importantly, this procedure efficiently determines the number of relevant features given an initial starting point. We provide freely available software implementations of the proposed methodology.
Introduction: Reductions of cerebrospinal fluid (CSF) amyloid-beta(Aβ42) and elevated phosphoryla... more Introduction: Reductions of cerebrospinal fluid (CSF) amyloid-beta(Aβ42) and elevated phosphorylated-tau(p-Tau) reflect in vivo Alzheimer's disease (AD) pathology and show utility in predicting conversion from mild cognitive impairment (MCI) to dementia. We investigated the P50 event related potential component as a non invasive biomarker of AD pathology in non-demented elderly. Methods: 36 MCI patients were stratified into amyloid positive (MCI-AD, n=17) and negative (MCI-Other, n=19) groups using CSF levels of Aβ42. All amyloid positive patients were also p-Tau positive. P50s were elicited with an auditory oddball paradigm. Results: MCI-AD patients yielded larger P50s than MCI-Other. The best amyloid-statuspredictor model showed 94.7% sensitivity, 94.1% specificity and 94.4% total accuracy. Discussion: P50 predicted amyloid status in MCI patients, thereby showing a relationship with AD pathology versus MCI from an other etiology. The P50 may have clinical utility for inexpensive prescreening and assessment of Alzheimer's pathology.
—Recent advances in machine learning, specifically in deep learning with neural networks, has mad... more —Recent advances in machine learning, specifically in deep learning with neural networks, has made a profound impact on fields such as natural language processing, image classification, and language modeling; however, feasibility and potential benefits of the approaches to metagenomic data analysis has been largely under-explored. Deep learning exploits many layers of learning nonlinear feature representations, typically in an unsupervised fashion, and recent results have shown outstanding generalization performance on previously unseen data. Furthermore, some deep learning methods can also represent the structure in a data set. Consequently, deep learning and neural networks may prove to be an appropriate approach for metagenomic data. To determine whether such approaches are indeed appropriate for metage-nomics, we experiment with two deep learning methods: i) a deep belief network, and ii) a recursive neural network, the latter of which provides a tree representing the structure of the data. We compare these approaches to the standard multi-layer perceptron, which has been well-established in the machine learning community as a powerful prediction algorithm, though its presence is largely missing in metagenomics literature. We find that traditional neural networks can be quite powerful classifiers on metagenomic data compared to baseline methods, such as random forests. On the other hand, while the deep learning approaches did not result in improvements to the classification accuracy, they do provide the ability to learn hierarchical representations of a data set that standard classification methods do not allow. Our goal in this effort is not to determine the best algorithm in terms accuracy—as that depends on the specific application—but rather to highlight the benefits and drawbacks of each of the approach we discuss and provide insight on how they can be improved for predictive metagenomic analysis.
— An increasing number of real-world applications are associated with streaming data drawn from d... more — An increasing number of real-world applications are associated with streaming data drawn from drifting and nonstationary distributions that change over time. These applications demand new algorithms that can learn and adapt to such changes, also known as concept drift. Proper characterization of such data with existing approaches typically requires substantial amount of labeled instances, which may be difficult, expensive, or even impractical to obtain. In this paper, we introduce compacted object sample extraction (COMPOSE), a computational geometry-based framework to learn from nonstationary streaming data, where labels are unavailable (or presented very sporadically) after initialization. We introduce the algorithm in detail, and discuss its results and performances on several synthetic and real-world data sets, which demonstrate the ability of the algorithm to learn under several different scenarios of initially labeled streaming environments. On carefully designed synthetic data sets, we compare the performance of COMPOSE against the optimal Bayes classifier, as well as the arbitrary subpopula-tion tracker algorithm, which addresses a similar environment referred to as extreme verification latency. Furthermore, using the real-world National Oceanic and Atmospheric Administration weather data set, we demonstrate that COMPOSE is competitive even with a well-established and fully supervised nonstationary learning algorithm that receives labeled data in every batch. Index Terms— Alpha shape, concept drift, nonstationary environment, semisupervised learning (SSL), verification latency.
Applications that generate data from
nonstationary environments, where the underlying
phenomena c... more Applications that generate data from nonstationary environments, where the underlying phenomena change over time, are becoming increasingly prevalent. Examples of these applications include making inferences or predictions based on financial data, energy demand and climate data analysis, web usage or sensor network monitoring, and malware/spam detection, among many others. In nonstationary environments, particularly those that generate streaming or multi-domain data, the probability density function of the data-generating process may change (drift) over time. Therefore, the fundamental and rather naïve assumption made by most computational intelligence approaches – that the training and testing data are sampled from the same fixed, albeit unknown, probability distribution – is simply not true. Learning in nonstationary environments requires adaptive or evolving approaches that can monitor and track the underlying changes, and adapt a model to accommodate those changes accordingly. In this effort, we provide a comprehensive survey and tutorial of established as well as state-of-the-art approaches, while highlighting two primary perspectives, active and passive, for learning in nonstationary environments. Finally, we also provide an inventory of existing real and synthetic datasets, as well as tools and software for getting started, evaluating and comparing different approaches.
IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS SPEECH AND SIGNAL …, Jan 1, 2002
We introduce a supervised learning algorithm that gives neural network classification algorithms ... more We introduce a supervised learning algorithm that gives neural network classification algorithms the capability of leaming incrementally from new data without forgetting what has been learned in earlier training sessions. Schapire's boosting algorithm, originally intended for improving the accuracy of weak leamers, has been modified to be used in an incremental leaming setting. The algorithm is based on generating a number of hypotheses using different distributions of the training data and combining these hypotheses using a weighted majority voting. This scheme allows the classifier previously trained with a training database, to leam from new data when the original data is no longer available, even when new classes are introduced. Initial results on incremental training of multilayer perceptron networks on synthetic as well as real-world data are presented in this paper.
To keep up with rapidly advancing technology, numerous innovations to the ECE curriculum, learnin... more To keep up with rapidly advancing technology, numerous innovations to the ECE curriculum, learning methods and pedagogy have been tested, implemented and envisioned. It is safe to say that no single approach will work for all of the diverse ECE technologies and every type of learner. However, a few key innovations appear useful in keeping undergraduate students motivated to learn, resilient to technology evolution and oriented amidst the overload of new information and ECE applications. Engineering clinics, similar to their medical clinic counterparts, provide project-based experiences within the core of an ECE education that enable transformation of the entire curriculum toward an outcomes-oriented, student centered, total quality environment. Clinics and project based learning approaches build skills within the individuals that give them confidence and motivation to continuously self-learn and adapt as the technologies around them give way to new, more effective paradigms. Perhaps more importantly engineering clinic experiences provide numerous opportunities for students to experience the holism of true engineering problem solving approaches and ranges of potential technology solutions. This paper reviews many of the innovations that will enable ECE education to become more effective in the midst of our present plethora of information and technology. Specific benefits from published and unpublished findings from engineering clinic and project-based learning experiences in actual use (Olin, Harvey Mudd, FIT, Drexel, Rose-Holman and Rowan) are summarized and discussed. This paper concludes that creating agile learning environments to graduate engineers that can be rapidly productive in the professional and research worlds is at least enhanced by some degree of clinic and/or project based learning experiences in the ECE curriculum.
Using a computational model to learn under various environments has been a well-researched field ... more Using a computational model to learn under various environments has been a well-researched field that produced relevant results; unfortunately, the majority of these efforts rely on three fundamental assumptions: i) there is a suffi-cient and representative data set to configure and assess ...
... on the specific class and the complexity of the particular laboratory exercise, the ... a bro... more ... on the specific class and the complexity of the particular laboratory exercise, the ... a broad background in biomedical engineering, and it is not just conducting random experiments ... In allexperiments described below (unless noted otherwise) students acquired their own biological ...
Alzheimer's disease (AD) is a neurodegenerative disease whose definitive diagnosis is only possib... more Alzheimer's disease (AD) is a neurodegenerative disease whose definitive diagnosis is only possible via autopsy. Currently used diagnostic approaches include the traditional neuropsychological tests, and recently more objective biomarkers, such as those obtained from cerebral spinal fluid (CSF), magnetic imaging resonance (MRI), and positron emission tomography (PET). Electroencephalography (EEG), a lower cost and noninvasive alternative, has been previously tried but with mixed success. In this effort, we attempt a more comprehensive analysis and comparison of machine learning approaches using EEG based features to determine diagnostic utility of the EEG. We compared support vector machine (SVM), naïve Bayes, multilayer perceptron (MLP), CART trees, k-nearest neighbor (kNN), and AdaBoost on various sets of features extracted from event related potentials (ERP) of the EEG. Our analysis suggests that there is indeed diagnostically useful information in the ERP of the EEG for early diagnosis of AD.
The ability to learn incrementally from streaming data -either in an online or batch setting -is ... more The ability to learn incrementally from streaming data -either in an online or batch setting -is of crucial importance for a prediction algorithm to learn from environments that generate vast amounts of data, where it is impractical or simply unfeasible to store all historical data. On the other hand, learning from streaming data becomes increasingly difficult when the probability distribution generating the data stream evolves over time, which renders the classification model generated from previously seen data suboptimal or potentially useless. Ensemble systems that employ multiple classifiers may be used to mitigate this effect, but even in such cases some classifiers (experts) become less knowledgeable for predicting on different domains than others as the distribution drifts. Further complication results when labeled data from a prediction (target) domain is not immediately available; hence, causing prediction on the target domain to yield sub-optimal results. In this work, we provide upper bounds on the loss, which hold with high probability, of a multiple expert system trained in such a nonstationary environment with verification latency. Furthermore, we show why a single model selection strategy can lead to undesirable results when learning in such nonstationary streaming settings. We present our analytical results with experiments on simulated as well as real-world data sets, comparing several different ensemble approaches to a single model.
A novel method of classifying data drawn from a nonstationary distribution with drifting mean and... more A novel method of classifying data drawn from a nonstationary distribution with drifting mean and variance is presented. The novelty of the approach is based on splitting the problem of tracking a nonstationary distribution into separate classification and time series state estimation problems. State space models for drift in both the mean and variance are presented, which are then successfully tracked using a Kalman filter and a particle filter for the linear and non-linear parts respectively. Preliminary results, which show the promising potential of the approach, are also presented, along with concluding remarks for potential uses of the proposed approach.
Learning in nonstationary environments, also called concept drift, requires an algorithm to track... more Learning in nonstationary environments, also called concept drift, requires an algorithm to track and learn from streaming data, drawn from a nonstationary (drifting) distribution. When data arrive continuously, a concept drift algorithm is required to maintain an up-to-date hypothesis that evolves with the changing environment. A more difficult problem that has received less attention, however, is learning from socalled initially labeled nonstationary environments, where the the environment provides only unlabeled data after initialization. Since the labels to such data never become available, learning in such a setting is also referred to as extreme verification latency, where the algorithm must only use unlabeled data to keep the hypothesis current. In this contribution, we analyze COMPOSE, a framework recently proposed for learning in such environments. One of the central processes of COMPOSE is core support extraction, where the algorithm predicts which data instances will be useful and relevant for classification in future time steps. We compare two different options, namely Gaussian mixture model based maximum a posteriori sampling and αshape compaction, for core support extraction, and analyze their effects on both accuracy and computational complexity of the algorithm. Our findings point to-as is the case in most engineering problems-a trade-off: that α-shapes are more versatile in most situations, but they are far more computationally complex, especially as the dimensionality of the dataset increases. Our proposed GMM procedure allows COMPOSE to operate on datasets of substantially larger dimensionality without affecting its classification performance.
The 2013 International Joint Conference on Neural Networks (IJCNN), 2013
ABSTRACT Increasing number of practical applications that involve streaming nonstationary data ha... more ABSTRACT Increasing number of practical applications that involve streaming nonstationary data have led to a recent surge in algorithms designed to learn from such data. One challenging version of this problem that has not received as much attention, however, is learning streaming nonstationary data when a small initial set of data are labeled, with unlabeled data being available thereafter. We have recently introduced the COMPOSE algorithm for learning in such scenarios, which we refer to as initially labeled nonstationary streaming data. COMPOSE works remarkably well, however it requires limited (gradual) drift, and cannot address special cases such as introduction of a new class or significant overlap of existing classes, as such scenarios cannot be learned without additional labeled data. Scenarios that provide occasional or periodic limited labeled data are not uncommon, however, for which many of COMPOSE's restrictions can be lifted. In this contribution, we describe a new version of COMPOSE as a proof-of-concept algorithm that can identify the instances whose labels — if available — would be most beneficial, and then combine those instances with unlabeled data to actively learn from streaming nonstationary data, even when the distribution of the data experiences abrupt changes. On two carefully designed experiments that include abrupt changes as well as addition of new classes, we show that COMPOSE.AL significantly outperforms original COMPOSE, while closely tracking the optimal Bayes classifier performance.
— Selection of most informative features that leads to a small loss on future data are arguably o... more — Selection of most informative features that leads to a small loss on future data are arguably one of the most important steps in classification, data analysis and model selection. Several feature selection (FS) algorithms are available; however, due to noise present in any data set, FS algorithms are typically accompanied by an appropriate cross-validation scheme. In this brief, we propose a statistical hypothesis test derived from the Neyman–Pearson lemma for determining if a feature is statistically relevant. The proposed approach can be applied as a wrapper to any FS algorithm, regardless of the FS criteria used by that algorithm, to determine whether a feature belongs in the relevant set. Perhaps more importantly, this procedure efficiently determines the number of relevant features given an initial starting point. We provide freely available software implementations of the proposed methodology.
Introduction: Reductions of cerebrospinal fluid (CSF) amyloid-beta(Aβ42) and elevated phosphoryla... more Introduction: Reductions of cerebrospinal fluid (CSF) amyloid-beta(Aβ42) and elevated phosphorylated-tau(p-Tau) reflect in vivo Alzheimer's disease (AD) pathology and show utility in predicting conversion from mild cognitive impairment (MCI) to dementia. We investigated the P50 event related potential component as a non invasive biomarker of AD pathology in non-demented elderly. Methods: 36 MCI patients were stratified into amyloid positive (MCI-AD, n=17) and negative (MCI-Other, n=19) groups using CSF levels of Aβ42. All amyloid positive patients were also p-Tau positive. P50s were elicited with an auditory oddball paradigm. Results: MCI-AD patients yielded larger P50s than MCI-Other. The best amyloid-statuspredictor model showed 94.7% sensitivity, 94.1% specificity and 94.4% total accuracy. Discussion: P50 predicted amyloid status in MCI patients, thereby showing a relationship with AD pathology versus MCI from an other etiology. The P50 may have clinical utility for inexpensive prescreening and assessment of Alzheimer's pathology.
—Recent advances in machine learning, specifically in deep learning with neural networks, has mad... more —Recent advances in machine learning, specifically in deep learning with neural networks, has made a profound impact on fields such as natural language processing, image classification, and language modeling; however, feasibility and potential benefits of the approaches to metagenomic data analysis has been largely under-explored. Deep learning exploits many layers of learning nonlinear feature representations, typically in an unsupervised fashion, and recent results have shown outstanding generalization performance on previously unseen data. Furthermore, some deep learning methods can also represent the structure in a data set. Consequently, deep learning and neural networks may prove to be an appropriate approach for metagenomic data. To determine whether such approaches are indeed appropriate for metage-nomics, we experiment with two deep learning methods: i) a deep belief network, and ii) a recursive neural network, the latter of which provides a tree representing the structure of the data. We compare these approaches to the standard multi-layer perceptron, which has been well-established in the machine learning community as a powerful prediction algorithm, though its presence is largely missing in metagenomics literature. We find that traditional neural networks can be quite powerful classifiers on metagenomic data compared to baseline methods, such as random forests. On the other hand, while the deep learning approaches did not result in improvements to the classification accuracy, they do provide the ability to learn hierarchical representations of a data set that standard classification methods do not allow. Our goal in this effort is not to determine the best algorithm in terms accuracy—as that depends on the specific application—but rather to highlight the benefits and drawbacks of each of the approach we discuss and provide insight on how they can be improved for predictive metagenomic analysis.
— An increasing number of real-world applications are associated with streaming data drawn from d... more — An increasing number of real-world applications are associated with streaming data drawn from drifting and nonstationary distributions that change over time. These applications demand new algorithms that can learn and adapt to such changes, also known as concept drift. Proper characterization of such data with existing approaches typically requires substantial amount of labeled instances, which may be difficult, expensive, or even impractical to obtain. In this paper, we introduce compacted object sample extraction (COMPOSE), a computational geometry-based framework to learn from nonstationary streaming data, where labels are unavailable (or presented very sporadically) after initialization. We introduce the algorithm in detail, and discuss its results and performances on several synthetic and real-world data sets, which demonstrate the ability of the algorithm to learn under several different scenarios of initially labeled streaming environments. On carefully designed synthetic data sets, we compare the performance of COMPOSE against the optimal Bayes classifier, as well as the arbitrary subpopula-tion tracker algorithm, which addresses a similar environment referred to as extreme verification latency. Furthermore, using the real-world National Oceanic and Atmospheric Administration weather data set, we demonstrate that COMPOSE is competitive even with a well-established and fully supervised nonstationary learning algorithm that receives labeled data in every batch. Index Terms— Alpha shape, concept drift, nonstationary environment, semisupervised learning (SSL), verification latency.
Applications that generate data from
nonstationary environments, where the underlying
phenomena c... more Applications that generate data from nonstationary environments, where the underlying phenomena change over time, are becoming increasingly prevalent. Examples of these applications include making inferences or predictions based on financial data, energy demand and climate data analysis, web usage or sensor network monitoring, and malware/spam detection, among many others. In nonstationary environments, particularly those that generate streaming or multi-domain data, the probability density function of the data-generating process may change (drift) over time. Therefore, the fundamental and rather naïve assumption made by most computational intelligence approaches – that the training and testing data are sampled from the same fixed, albeit unknown, probability distribution – is simply not true. Learning in nonstationary environments requires adaptive or evolving approaches that can monitor and track the underlying changes, and adapt a model to accommodate those changes accordingly. In this effort, we provide a comprehensive survey and tutorial of established as well as state-of-the-art approaches, while highlighting two primary perspectives, active and passive, for learning in nonstationary environments. Finally, we also provide an inventory of existing real and synthetic datasets, as well as tools and software for getting started, evaluating and comparing different approaches.
IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS SPEECH AND SIGNAL …, Jan 1, 2002
We introduce a supervised learning algorithm that gives neural network classification algorithms ... more We introduce a supervised learning algorithm that gives neural network classification algorithms the capability of leaming incrementally from new data without forgetting what has been learned in earlier training sessions. Schapire's boosting algorithm, originally intended for improving the accuracy of weak leamers, has been modified to be used in an incremental leaming setting. The algorithm is based on generating a number of hypotheses using different distributions of the training data and combining these hypotheses using a weighted majority voting. This scheme allows the classifier previously trained with a training database, to leam from new data when the original data is no longer available, even when new classes are introduced. Initial results on incremental training of multilayer perceptron networks on synthetic as well as real-world data are presented in this paper.
To keep up with rapidly advancing technology, numerous innovations to the ECE curriculum, learnin... more To keep up with rapidly advancing technology, numerous innovations to the ECE curriculum, learning methods and pedagogy have been tested, implemented and envisioned. It is safe to say that no single approach will work for all of the diverse ECE technologies and every type of learner. However, a few key innovations appear useful in keeping undergraduate students motivated to learn, resilient to technology evolution and oriented amidst the overload of new information and ECE applications. Engineering clinics, similar to their medical clinic counterparts, provide project-based experiences within the core of an ECE education that enable transformation of the entire curriculum toward an outcomes-oriented, student centered, total quality environment. Clinics and project based learning approaches build skills within the individuals that give them confidence and motivation to continuously self-learn and adapt as the technologies around them give way to new, more effective paradigms. Perhaps more importantly engineering clinic experiences provide numerous opportunities for students to experience the holism of true engineering problem solving approaches and ranges of potential technology solutions. This paper reviews many of the innovations that will enable ECE education to become more effective in the midst of our present plethora of information and technology. Specific benefits from published and unpublished findings from engineering clinic and project-based learning experiences in actual use (Olin, Harvey Mudd, FIT, Drexel, Rose-Holman and Rowan) are summarized and discussed. This paper concludes that creating agile learning environments to graduate engineers that can be rapidly productive in the professional and research worlds is at least enhanced by some degree of clinic and/or project based learning experiences in the ECE curriculum.
Using a computational model to learn under various environments has been a well-researched field ... more Using a computational model to learn under various environments has been a well-researched field that produced relevant results; unfortunately, the majority of these efforts rely on three fundamental assumptions: i) there is a suffi-cient and representative data set to configure and assess ...
... on the specific class and the complexity of the particular laboratory exercise, the ... a bro... more ... on the specific class and the complexity of the particular laboratory exercise, the ... a broad background in biomedical engineering, and it is not just conducting random experiments ... In allexperiments described below (unless noted otherwise) students acquired their own biological ...
Alzheimer's disease (AD) is a neurodegenerative disease whose definitive diagnosis is only possib... more Alzheimer's disease (AD) is a neurodegenerative disease whose definitive diagnosis is only possible via autopsy. Currently used diagnostic approaches include the traditional neuropsychological tests, and recently more objective biomarkers, such as those obtained from cerebral spinal fluid (CSF), magnetic imaging resonance (MRI), and positron emission tomography (PET). Electroencephalography (EEG), a lower cost and noninvasive alternative, has been previously tried but with mixed success. In this effort, we attempt a more comprehensive analysis and comparison of machine learning approaches using EEG based features to determine diagnostic utility of the EEG. We compared support vector machine (SVM), naïve Bayes, multilayer perceptron (MLP), CART trees, k-nearest neighbor (kNN), and AdaBoost on various sets of features extracted from event related potentials (ERP) of the EEG. Our analysis suggests that there is indeed diagnostically useful information in the ERP of the EEG for early diagnosis of AD.
The ability to learn incrementally from streaming data -either in an online or batch setting -is ... more The ability to learn incrementally from streaming data -either in an online or batch setting -is of crucial importance for a prediction algorithm to learn from environments that generate vast amounts of data, where it is impractical or simply unfeasible to store all historical data. On the other hand, learning from streaming data becomes increasingly difficult when the probability distribution generating the data stream evolves over time, which renders the classification model generated from previously seen data suboptimal or potentially useless. Ensemble systems that employ multiple classifiers may be used to mitigate this effect, but even in such cases some classifiers (experts) become less knowledgeable for predicting on different domains than others as the distribution drifts. Further complication results when labeled data from a prediction (target) domain is not immediately available; hence, causing prediction on the target domain to yield sub-optimal results. In this work, we provide upper bounds on the loss, which hold with high probability, of a multiple expert system trained in such a nonstationary environment with verification latency. Furthermore, we show why a single model selection strategy can lead to undesirable results when learning in such nonstationary streaming settings. We present our analytical results with experiments on simulated as well as real-world data sets, comparing several different ensemble approaches to a single model.
A novel method of classifying data drawn from a nonstationary distribution with drifting mean and... more A novel method of classifying data drawn from a nonstationary distribution with drifting mean and variance is presented. The novelty of the approach is based on splitting the problem of tracking a nonstationary distribution into separate classification and time series state estimation problems. State space models for drift in both the mean and variance are presented, which are then successfully tracked using a Kalman filter and a particle filter for the linear and non-linear parts respectively. Preliminary results, which show the promising potential of the approach, are also presented, along with concluding remarks for potential uses of the proposed approach.
Learning in nonstationary environments, also called concept drift, requires an algorithm to track... more Learning in nonstationary environments, also called concept drift, requires an algorithm to track and learn from streaming data, drawn from a nonstationary (drifting) distribution. When data arrive continuously, a concept drift algorithm is required to maintain an up-to-date hypothesis that evolves with the changing environment. A more difficult problem that has received less attention, however, is learning from socalled initially labeled nonstationary environments, where the the environment provides only unlabeled data after initialization. Since the labels to such data never become available, learning in such a setting is also referred to as extreme verification latency, where the algorithm must only use unlabeled data to keep the hypothesis current. In this contribution, we analyze COMPOSE, a framework recently proposed for learning in such environments. One of the central processes of COMPOSE is core support extraction, where the algorithm predicts which data instances will be useful and relevant for classification in future time steps. We compare two different options, namely Gaussian mixture model based maximum a posteriori sampling and αshape compaction, for core support extraction, and analyze their effects on both accuracy and computational complexity of the algorithm. Our findings point to-as is the case in most engineering problems-a trade-off: that α-shapes are more versatile in most situations, but they are far more computationally complex, especially as the dimensionality of the dataset increases. Our proposed GMM procedure allows COMPOSE to operate on datasets of substantially larger dimensionality without affecting its classification performance.
Uploads
Papers by Robi Polikar
Methods: 36 MCI patients were stratified into amyloid positive (MCI-AD, n=17) and negative (MCI-Other, n=19) groups using CSF levels of Aβ42. All amyloid positive patients were also
p-Tau positive. P50s were elicited with an auditory oddball paradigm.
Results: MCI-AD patients yielded larger P50s than MCI-Other. The best amyloid-statuspredictor model showed 94.7% sensitivity, 94.1% specificity and 94.4% total accuracy.
Discussion: P50 predicted amyloid status in MCI patients, thereby showing a relationship with AD pathology versus MCI from an other etiology. The P50 may have clinical utility for
inexpensive prescreening and assessment of Alzheimer's pathology.
nonstationary environments, where the underlying
phenomena change over time, are becoming increasingly
prevalent. Examples of these applications include making
inferences or predictions based on financial data, energy demand
and climate data analysis, web usage or sensor network monitoring, and
malware/spam detection, among many others. In nonstationary environments,
particularly those that generate streaming or multi-domain data, the probability
density function of the data-generating process may change (drift) over time.
Therefore, the fundamental and rather naïve assumption made by most computational
intelligence approaches – that the training and testing data are sampled from the same fixed,
albeit unknown, probability distribution – is simply not true. Learning in nonstationary
environments requires adaptive or evolving approaches that can monitor and track the
underlying changes, and adapt a model to accommodate those changes accordingly.
In this effort, we provide a comprehensive survey and tutorial of established as
well as state-of-the-art approaches, while highlighting two primary perspectives,
active and passive, for learning in nonstationary environments.
Finally, we also provide an inventory of existing real and
synthetic datasets, as well as tools and software for getting
started, evaluating and comparing different approaches.
Methods: 36 MCI patients were stratified into amyloid positive (MCI-AD, n=17) and negative (MCI-Other, n=19) groups using CSF levels of Aβ42. All amyloid positive patients were also
p-Tau positive. P50s were elicited with an auditory oddball paradigm.
Results: MCI-AD patients yielded larger P50s than MCI-Other. The best amyloid-statuspredictor model showed 94.7% sensitivity, 94.1% specificity and 94.4% total accuracy.
Discussion: P50 predicted amyloid status in MCI patients, thereby showing a relationship with AD pathology versus MCI from an other etiology. The P50 may have clinical utility for
inexpensive prescreening and assessment of Alzheimer's pathology.
nonstationary environments, where the underlying
phenomena change over time, are becoming increasingly
prevalent. Examples of these applications include making
inferences or predictions based on financial data, energy demand
and climate data analysis, web usage or sensor network monitoring, and
malware/spam detection, among many others. In nonstationary environments,
particularly those that generate streaming or multi-domain data, the probability
density function of the data-generating process may change (drift) over time.
Therefore, the fundamental and rather naïve assumption made by most computational
intelligence approaches – that the training and testing data are sampled from the same fixed,
albeit unknown, probability distribution – is simply not true. Learning in nonstationary
environments requires adaptive or evolving approaches that can monitor and track the
underlying changes, and adapt a model to accommodate those changes accordingly.
In this effort, we provide a comprehensive survey and tutorial of established as
well as state-of-the-art approaches, while highlighting two primary perspectives,
active and passive, for learning in nonstationary environments.
Finally, we also provide an inventory of existing real and
synthetic datasets, as well as tools and software for getting
started, evaluating and comparing different approaches.