Effectiveness of document representation for classification

Conventionally, document classification researches focus on improving the learning capabilities of classifiers. Nevertheless, according to our observation, the effectiveness of classification is limited by the suitability of document representation. Intuitively, the more features that are used in representation, the more comprehensive that documents are represented. However, if a representation contains too many irrelevant features, the classifier would suffer from not only the curse of high dimensionality, but also overfitting. To address this problem of suitableness of document representations, we present a classifier-independent approach to measure the effectiveness of document representations. Our approach utilises a labelled document corpus to estimate the distribution of documents in the feature space. By looking through documents in this way, we can clearly identify the contributions made by different features toward the document classification. Some experiments have been performed to show how the effectiveness is evaluated. Our approach can be used as a tool to assist feature selection, dimensionality reduction and document classification.

Using Bayesian neural networks to classify segmented images

We present results that compare the performance of neural networks trained with two Bayesian methods, (i) the Evidence Framework of MacKay (1992) and (ii) a Markov Chain Monte Carlo method due to Neal (1996) on a task of classifying segmented outdoor images. We also investigate the use of the Automatic Relevance Determination method for input feature selection.

EM algorithm for GTM-FS

We propose a generative topographic mapping (GTM) based data visualization with simultaneous feature selection (GTM-FS) approach which not only provides a better visualization by modeling irrelevant features ("noise") using a separate shared distribution but also gives a saliency value for each feature which helps the user to assess their significance. This technical report presents a varient of the Expectation-Maximization (EM) algorithm for GTM-FS.

A study of cluster analysis techniques and their applications

This thesis seeks to describe the development of an inexpensive and efficient clustering technique for multivariate data analysis. The technique starts from a multivariate data matrix and ends with graphical representation of the data and pattern recognition discriminant function. The technique also results in distances frequency distribution that might be useful in detecting clustering in the data or for the estimation of parameters useful in the discrimination between the different populations in the data. The technique can also be used in feature selection. The technique is essentially for the discovery of data structure by revealing the component parts of the data. lhe thesis offers three distinct contributions for cluster analysis and pattern recognition techniques. The first contribution is the introduction of transformation function in the technique of nonlinear mapping. The second contribution is the us~ of distances frequency distribution instead of distances time-sequence in nonlinear mapping, The third contribution is the formulation of a new generalised and normalised error function together with its optimal step size formula for gradient method minimisation. The thesis consists of five chapters. The first chapter is the introduction. The second chapter describes multidimensional scaling as an origin of nonlinear mapping technique. The third chapter describes the first developing step in the technique of nonlinear mapping that is the introduction of "transformation function". The fourth chapter describes the second developing step of the nonlinear mapping technique. This is the use of distances frequency distribution instead of distances time-sequence. The chapter also includes the new generalised and normalised error function formulation. Finally, the fifth chapter, the conclusion, evaluates all developments and proposes a new program. for cluster analysis and pattern recognition by integrating all the new features.

Novel speech signal processing algorithms for high-accuracy classification of Parkinson's disease

There has been considerable recent research into the connection between Parkinson's disease (PD) and speech impairment. Recently, a wide range of speech signal processing algorithms (dysphonia measures) aiming to predict PD symptom severity using speech signals have been introduced. In this paper, we test how accurately these novel algorithms can be used to discriminate PD subjects from healthy controls. In total, we compute 132 dysphonia measures from sustained vowels. Then, we select four parsimonious subsets of these dysphonia measures using four feature selection algorithms, and map these feature subsets to a binary classification response using two statistical classifiers: random forests and support vector machines. We use an existing database consisting of 263 samples from 43 subjects, and demonstrate that these new dysphonia measures can outperform state-of-the-art results, reaching almost 99% overall classification accuracy using only ten dysphonia features. We find that some of the recently proposed dysphonia measures complement existing algorithms in maximizing the ability of the classifiers to discriminate healthy controls from PD subjects. We see these results as an important step toward noninvasive diagnostic decision support in PD.

Nonlinear speech analysis algorithms mapped to a standard metric achieve clinically useful quantification of average Parkinson's disease symptom severity

The standard reference clinical score quantifying average Parkinson's disease (PD) symptom severity is the Unified Parkinson's Disease Rating Scale (UPDRS). At present, UPDRS is determined by the subjective clinical evaluation of the patient's ability to adequately cope with a range of tasks. In this study, we extend recent findings that UPDRS can be objectively assessed to clinically useful accuracy using simple, self-administered speech tests, without requiring the patient's physical presence in the clinic. We apply a wide range of known speech signal processing algorithms to a large database (approx. 6000 recordings from 42 PD patients, recruited to a six-month, multi-centre trial) and propose a number of novel, nonlinear signal processing algorithms which reveal pathological characteristics in PD more accurately than existing approaches. Robust feature selection algorithms select the optimal subset of these algorithms, which is fed into non-parametric regression and classification algorithms, mapping the signal processing algorithm outputs to UPDRS. We demonstrate rapid, accurate replication of the UPDRS assessment with clinically useful accuracy (about 2 UPDRS points difference from the clinicians' estimates, p < 0.001). This study supports the viability of frequent, remote, cost-effective, objective, accurate UPDRS telemonitoring based on self-administered speech tests. This technology could facilitate large-scale clinical trials into novel PD treatments.

Handling varying amounts of missing data when classifying mental-health risk levels

One of the main challenges of classifying clinical data is determining how to handle missing features. Most research favours imputing of missing values or neglecting records that include missing data, both of which can degrade accuracy when missing values exceed a certain level. In this research we propose a methodology to handle data sets with a large percentage of missing values and with high variability in which particular data are missing. Feature selection is effected by picking variables sequentially in order of maximum correlation with the dependent variable and minimum correlation with variables already selected. Classification models are generated individually for each test case based on its particular feature set and the matching data values available in the training population. The method was applied to real patients' anonymous mental-health data where the task was to predict the suicide risk judgement clinicians would give for each patient's data, with eleven possible outcome classes: zero to ten, representing no risk to maximum risk. The results compare favourably with alternative methods and have the advantage of ensuring explanations of risk are based only on the data given, not imputed data. This is important for clinical decision support systems using human expertise for modelling and explaining predictions.

Handling high level of censoring for endovascular aortic repair risk prediction

Feature selection is important in medical field for many reasons. However, selecting important variables is a difficult task with the presence of censoring that is a unique feature in survival data analysis. This paper proposed an approach to deal with the censoring problem in endovascular aortic repair survival data through Bayesian networks. It was merged and embedded with a hybrid feature selection process that combines cox's univariate analysis with machine learning approaches such as ensemble artificial neural networks to select the most relevant predictive variables. The proposed algorithm was compared with common survival variable selection approaches such as; least absolute shrinkage and selection operator LASSO, and Akaike information criterion AIC methods. The results showed that it was capable of dealing with high censoring in the datasets. Moreover, ensemble classifiers increased the area under the roc curves of the two datasets collected from two centers located in United Kingdom separately. Furthermore, ensembles constructed with center 1 enhanced the concordance index of center 2 prediction compared to the model built with a single network. Although the size of the final reduced model using the neural networks and its ensembles is greater than other methods, the model outperformed the others in both concordance index and sensitivity for center 2 prediction. This indicates the reduced model is more powerful for cross center prediction.

The risk of re-intervention after endovascular aortic aneurysm repair

This thesis studies survival analysis techniques dealing with censoring to produce predictive tools that predict the risk of endovascular aortic aneurysm repair (EVAR) re-intervention. Censoring indicates that some patients do not continue follow up, so their outcome class is unknown. Methods dealing with censoring have drawbacks and cannot handle the high censoring of the two EVAR datasets collected. Therefore, this thesis presents a new solution to high censoring by modifying an approach that was incapable of differentiating between risks groups of aortic complications. Feature selection (FS) becomes complicated with censoring. Most survival FS methods depends on Cox's model, however machine learning classifiers (MLC) are preferred. Few methods adopted MLC to perform survival FS, but they cannot be used with high censoring. This thesis proposes two FS methods which use MLC to evaluate features. The two FS methods use the new solution to deal with censoring. They combine factor analysis with greedy stepwise FS search which allows eliminated features to enter the FS process. The first FS method searches for the best neural networks' configuration and subset of features. The second approach combines support vector machines, neural networks, and K nearest neighbor classifiers using simple and weighted majority voting to construct a multiple classifier system (MCS) for improving the performance of individual classifiers. It presents a new hybrid FS process by using MCS as a wrapper method and merging it with the iterated feature ranking filter method to further reduce the features. The proposed techniques outperformed FS methods based on Cox's model such as; Akaike and Bayesian information criteria, and least absolute shrinkage and selector operator in the log-rank test's p-values, sensitivity, and concordance. This proves that the proposed techniques are more powerful in correctly predicting the risk of re-intervention. Consequently, they enable doctors to set patients’ appropriate future observation plan.

Ensemble fuzzy belief intrusion detection design

With the rapid growth of the Internet, computer attacks are increasing at a fast pace and can easily cause millions of dollar in damage to an organization. Detecting these attacks is an important issue of computer security. There are many types of attacks and they fall into four main categories, Denial of Service (DoS) attacks, Probe, User to Root (U2R) attacks, and Remote to Local (R2L) attacks. Within these categories, DoS and Probe attacks continuously show up with greater frequency in a short period of time when they attack systems. They are different from the normal traffic data and can be easily separated from normal activities. On the contrary, U2R and R2L attacks are embedded in the data portions of the packets and normally involve only a single connection. It becomes difficult to achieve satisfactory detection accuracy for detecting these two attacks. Therefore, we focus on studying the ambiguity problem between normal activities and U2R/R2L attacks. The goal is to build a detection system that can accurately and quickly detect these two attacks. In this dissertation, we design a two-phase intrusion detection approach. In the first phase, a correlation-based feature selection algorithm is proposed to advance the speed of detection. Features with poor prediction ability for the signatures of attacks and features inter-correlated with one or more other features are considered redundant. Such features are removed and only indispensable information about the original feature space remains. In the second phase, we develop an ensemble intrusion detection system to achieve accurate detection performance. The proposed method includes multiple feature selecting intrusion detectors and a data mining intrusion detector. The former ones consist of a set of detectors, and each of them uses a fuzzy clustering technique and belief theory to solve the ambiguity problem. The latter one applies data mining technique to automatically extract computer users’ normal behavior from training network traffic data. The final decision is a combination of the outputs of feature selecting and data mining detectors. The experimental results indicate that our ensemble approach not only significantly reduces the detection time but also effectively detect U2R and R2L attacks that contain degrees of ambiguous information.

Exploring the integration of data mining and data visualization

Due to the rapid advances in computing and sensing technologies, enormous amounts of data are being generated everyday in various applications. The integration of data mining and data visualization has been widely used to analyze these massive and complex data sets to discover hidden patterns. For both data mining and visualization to be effective, it is important to include the visualization techniques in the mining process and to generate the discovered patterns for a more comprehensive visual view. In this dissertation, four related problems: dimensionality reduction for visualizing high dimensional datasets, visualization-based clustering evaluation, interactive document mining, and multiple clusterings exploration are studied to explore the integration of data mining and data visualization. In particular, we 1) propose an efficient feature selection method (reliefF + mRMR) for preprocessing high dimensional datasets; 2) present DClusterE to integrate cluster validation with user interaction and provide rich visualization tools for users to examine document clustering results from multiple perspectives; 3) design two interactive document summarization systems to involve users efforts and generate customized summaries from 2D sentence layouts; and 4) propose a new framework which organizes the different input clusterings into a hierarchical tree structure and allows for interactive exploration of multiple clustering solutions.

Ensemble Fuzzy Belief Intrusion Detection Design

With the rapid growth of the Internet, computer attacks are increasing at a fast pace and can easily cause millions of dollar in damage to an organization. Detecting these attacks is an important issue of computer security. There are many types of attacks and they fall into four main categories, Denial of Service (DoS) attacks, Probe, User to Root (U2R) attacks, and Remote to Local (R2L) attacks. Within these categories, DoS and Probe attacks continuously show up with greater frequency in a short period of time when they attack systems. They are different from the normal traffic data and can be easily separated from normal activities. On the contrary, U2R and R2L attacks are embedded in the data portions of the packets and normally involve only a single connection. It becomes difficult to achieve satisfactory detection accuracy for detecting these two attacks. Therefore, we focus on studying the ambiguity problem between normal activities and U2R/R2L attacks. The goal is to build a detection system that can accurately and quickly detect these two attacks. In this dissertation, we design a two-phase intrusion detection approach. In the first phase, a correlation-based feature selection algorithm is proposed to advance the speed of detection. Features with poor prediction ability for the signatures of attacks and features inter-correlated with one or more other features are considered redundant. Such features are removed and only indispensable information about the original feature space remains. In the second phase, we develop an ensemble intrusion detection system to achieve accurate detection performance. The proposed method includes multiple feature selecting intrusion detectors and a data mining intrusion detector. The former ones consist of a set of detectors, and each of them uses a fuzzy clustering technique and belief theory to solve the ambiguity problem. The latter one applies data mining technique to automatically extract computer users’ normal behavior from training network traffic data. The final decision is a combination of the outputs of feature selecting and data mining detectors. The experimental results indicate that our ensemble approach not only significantly reduces the detection time but also effectively detect U2R and R2L attacks that contain degrees of ambiguous information.

Classificação de atividade eletromiográfia facial de indivíduos saudáveis e com hanseníese por meio de máquina de vetores de suporte

A number of studies in the areas of Biomedical Engineering and Health Sciences have employed machine learning tools to develop methods capable of identifying patterns in different sets of data. Despite its extinction in many countries of the developed world, Hansen’s disease is still a disease that affects a huge part of the population in countries such as India and Brazil. In this context, this research proposes to develop a method that makes it possible to understand in the future how Hansen’s disease affects facial muscles. By using surface electromyography, a system was adapted so as to capture the signals from the largest possible number of facial muscles. We have first looked upon the literature to learn about the way researchers around the globe have been working with diseases that affect the peripheral neural system and how electromyography has acted to contribute to the understanding of these diseases. From these data, a protocol was proposed to collect facial surface electromyographic (sEMG) signals so that these signals presented a high signal to noise ratio. After collecting the signals, we looked for a method that would enable the visualization of this information in a way to make it possible to guarantee that the method used presented satisfactory results. After identifying the method's efficiency, we tried to understand which information could be extracted from the electromyographic signal representing the collected data. Once studies demonstrating which information could contribute to a better understanding of this pathology were not to be found in literature, parameters of amplitude, frequency and entropy were extracted from the signal and a feature selection was made in order to look for the features that better distinguish a healthy individual from a pathological one. After, we tried to identify the classifier that best discriminates distinct individuals from different groups, and also the set of parameters of this classifier that would bring the best outcome. It was identified that the protocol proposed in this study and the adaptation with disposable electrodes available in market proved their effectiveness and capability of being used in different studies whose intention is to collect data from facial electromyography. The feature selection algorithm also showed that not all of the features extracted from the signal are significant for data classification, with some more relevant than others. The classifier Support Vector Machine (SVM) proved itself efficient when the adequate Kernel function was used with the muscle from which information was to be extracted. Each investigated muscle presented different results when the classifier used linear, radial and polynomial kernel functions. Even though we have focused on Hansen’s disease, the method applied here can be used to study facial electromyography in other pathologies.

Classificação de atividade eletromiográfia facial de indivíduos saudáveis e com hanseníase por meio de máquina de vetores de suporte

A number of studies in the areas of Biomedical Engineering and Health Sciences have employed machine learning tools to develop methods capable of identifying patterns in different sets of data. Despite its extinction in many countries of the developed world, Hansen’s disease is still a disease that affects a huge part of the population in countries such as India and Brazil. In this context, this research proposes to develop a method that makes it possible to understand in the future how Hansen’s disease affects facial muscles. By using surface electromyography, a system was adapted so as to capture the signals from the largest possible number of facial muscles. We have first looked upon the literature to learn about the way researchers around the globe have been working with diseases that affect the peripheral neural system and how electromyography has acted to contribute to the understanding of these diseases. From these data, a protocol was proposed to collect facial surface electromyographic (sEMG) signals so that these signals presented a high signal to noise ratio. After collecting the signals, we looked for a method that would enable the visualization of this information in a way to make it possible to guarantee that the method used presented satisfactory results. After identifying the method's efficiency, we tried to understand which information could be extracted from the electromyographic signal representing the collected data. Once studies demonstrating which information could contribute to a better understanding of this pathology were not to be found in literature, parameters of amplitude, frequency and entropy were extracted from the signal and a feature selection was made in order to look for the features that better distinguish a healthy individual from a pathological one. After, we tried to identify the classifier that best discriminates distinct individuals from different groups, and also the set of parameters of this classifier that would bring the best outcome. It was identified that the protocol proposed in this study and the adaptation with disposable electrodes available in market proved their effectiveness and capability of being used in different studies whose intention is to collect data from facial electromyography. The feature selection algorithm also showed that not all of the features extracted from the signal are significant for data classification, with some more relevant than others. The classifier Support Vector Machine (SVM) proved itself efficient when the adequate Kernel function was used with the muscle from which information was to be extracted. Each investigated muscle presented different results when the classifier used linear, radial and polynomial kernel functions. Even though we have focused on Hansen’s disease, the method applied here can be used to study facial electromyography in other pathologies.

Computational Methods for Identifying Hypertrophic Cardiomyopathy using 12-lead ECG

Hypertrophic cardiomyopathy (HCM) is a cardiovascular disease where the heart muscle is partially thickened and blood flow is - potentially fatally - obstructed. It is one of the leading causes of sudden cardiac death in young people. Electrocardiography (ECG) and Echocardiography (Echo) are the standard tests for identifying HCM and other cardiac abnormalities. The American Heart Association has recommended using a pre-participation questionnaire for young athletes instead of ECG or Echo tests due to considerations of cost and time involved in interpreting the results of these tests by an expert cardiologist. Initially we set out to develop a classifier for automated prediction of young athletes’ heart conditions based on the answers to the questionnaire. Classification results and further in-depth analysis using computational and statistical methods indicated significant shortcomings of the questionnaire in predicting cardiac abnormalities. Automated methods for analyzing ECG signals can help reduce cost and save time in the pre-participation screening process by detecting HCM and other cardiac abnormalities. Therefore, the main goal of this dissertation work is to identify HCM through computational analysis of 12-lead ECG. ECG signals recorded on one or two leads have been analyzed in the past for classifying individual heartbeats into different types of arrhythmia as annotated primarily in the MIT-BIH database. In contrast, we classify complete sequences of 12-lead ECGs to assign patients into two groups: HCM vs. non-HCM. The challenges and issues we address include missing ECG waves in one or more leads and the dimensionality of a large feature-set. We address these by proposing imputation and feature-selection methods. We develop heartbeat-classifiers by employing Random Forests and Support Vector Machines, and propose a method to classify full 12-lead ECGs based on the proportion of heartbeats classified as HCM. The results from our experiments show that the classifiers developed using our methods perform well in identifying HCM. Thus the two contributions of this thesis are the utilization of computational and statistical methods for discovering shortcomings in a current screening procedure and the development of methods to identify HCM through computational analysis of 12-lead ECG signals.