Class decomposition for GA-based classifier agents--a Pitt approach.

This paper proposes a class decomposition approach to improve the performance of GA-based classifier agents. This approach partitions a classification problem into several class modules in the output domain, and each module is responsible for solving a fraction of the original problem. These modules are trained in parallel and independently, and results obtained from them are integrated to form the final solution by resolving conflicts. Benchmark classification data sets are used to evaluate the proposed approaches. The experiment results show that class decomposition can help achieve higher classification rate with training time reduced.     

Fusion of covariance matrices of PCA and FLD

In this paper, we propose a novel approach for fusing two classifiers, specifically classifiers based on subspace analysis, during feature extraction. A method of combining the covariance matrices of the Principal Component Analysis (PCA) and Fisher Linear Discriminant (FLD) is presented. Unlike other existing fusion strategies which fuse classifiers either at data level, or at feature level or at decision level, the proposed work combines two classifiers while extracting features introducing a new unexplored area for further research. The covariance matrices of PCA and FLD are combined using a product rule to preserve the natures of both covariance matrices with an expectation to have an increased performance. In order to show the effectiveness of the proposed fusion method, we have conducted a visual simulation on iris data. The proposed model has also been tested by performing clustering on standard datasets such as Zoo, Wine, and Iris. To study the versatility of the proposed method we have carried out an experimentation on sports video shot retrieval problem. The experimental results signify that the proposed fusing approach has an improved performance over individual classifiers.     

基于分解策略处理多分类不均衡问题的方法

针对多分类不均衡问题,提出了一种新的基于一对一（one-versus-one,OVO）分解策略的方法。首先基于OVO分解策略将多分类不均衡问题分解成多个二值分类问题;再利用处理不均衡二值分类问题的算法建立二值分类器;接着利用SMOTE过抽样技术处理原始数据集;然后采用基于距离相对竞争力加权方法处理冗余分类器;最后通过加权投票法获得输出结果。在KEEL不均衡数据集上的大量实验结果表明,所提算法比其他经典方法具有显著的优势。     

面向文本分类的混淆类判别技术

分析了文本分类过程中存在的混淆类现象,主要研究混淆类的判别技术,进而改善文本分类的性能.首先,提出了一种基于分类错误分布的混淆类识别技术,识别预定义类别中的混淆类集合.为了有效判别混淆类,提出了一种基于判别能力的特征选取技术,通过评价某一特征对类别之间的判别能力实现特征选取.最后,通过基于两阶段的分类器设计框架,将初始分类器和混淆类分类器进行集成,组合了两个阶段的分类结果作为最后输出.混淆类分类器的激活条件是:当测试文本被初始分类器标注为混淆类类别时,即采用混淆类分类器进行重新判别.在比较实验中采用了Newsgroup和863中文评测语料,针对单标签、多类分类器.实验结果显示,该技术有效地改善了分类性能.     

Comparing Combination Rules of Pairwise Neural Networks Classifiers

A decomposition approach to multiclass classification problems consists in decomposing a multiclass problem into a set of binary ones. Decomposition splits the complete multiclass problem into a set of smaller classification problems involving only two classes (binary classification: dichotomies). With a decomposition, one has to define a recombination which recomposes the outputs of the dichotomizers in order to solve the original multiclass problem. There are several approaches to the decomposition, the most famous ones being one-against-all and one-against-one also called pairwise. In this paper, we focus on pairwise decomposition approach to multiclass classification with neural networks as the base learner for the dichotomies. We are primarily interested in the different possible ways to perform the so-called recombination (or decoding). We review standard methods used to decode the decomposition generated by a one-against-one approach. New decoding methods are proposed and compared to standard methods. A stacking decoding is also proposed which consists in replacing the whole decoding or a part of it by a trainable classifier to arbiter among the conflicting predictions of the pairwise classifiers. Proposed methods try to cope with the main problem while using pairwise decomposition: the use of irrelevant classifiers. Substantial gain is obtained on all datasets used in the experiments. Based on the above, we provide future research directions which consider the recombination problem as an ensemble method.     

改进PM-MD的分类器集成

Hub会对高维数据分析产生显著消极影响,现有研究分别采用了五种降Hubness策略以提高分类效果,但单个降Hubness策略适用范围有限.为解决这一问题,提出对五种降Hub分类器进行基于PM-MD的集成,PM-MD集成是通过KNN确定分类对象的决策表以及通过分类器得到分类对象的类支持向量,最后通过比较决策表和类支持向量的相似性计算分类器的竞争力权重.由于PM-MD在处理高维数据集时高斯势函数存在弱化距离导致区分度不足的倾向,因此提出了采用欧氏距离直接计算决策表以提高分类精度.在12个UCI数据集上的实验结果表明：PM-MD不仅获得更好且稳定的分类结果,而改进后的PM-MD则进一步提高了分类精度.     

A Dynamic Spark-based Classification Framework for Imbalanced Big Data

Classification of imbalanced big data has assembled an extensive consideration by many researchers during the last decade. Standard classification methods poorly diagnosis the minority class samples. Several approaches have been introduced for solving the problem of class imbalance in big data to enhance the generalization in classification. However, most of these approaches neglect the effect of border samples on classification performance; the high impact border samples might expose to misclassification. In this paper, a Spark Based Mining Framework (SBMF) is proposed to address the imbalanced data problem. Two main modules are designed for this purpose. The first is the Border Handling Module (BHM) which under samples the low impact majority border instances and oversamples the minority class instances. The second module is the Selective Border Instances sampling (SBI) Module, which enhances the output of the BHM module. The performance of the SBMF framework is evaluated and compared with other recent systems. A number of experiments were performed using moderate and big datasets with different imbalanced ratio. The results obtained from SBMF framework, when compared to the recent works, show better performance for the different datasets and classifiers.     

Ensemble based on GA wrapper feature selection

Ensemble of classifiers is a learning paradigm where many classifiers are jointly used to solve a problem. Research has shown that ensemble is very effective for classification tasks. Diversity and accuracy are two basic requirements for the ensemble creation. In this paper, we propose an ensemble creation method based on GA wrapper feature selection. Preliminary experimental results on real-world data show that the proposed method is promising, especially when the number of training data is limited.     

Risk function estimation for subproblems in a hierarchical classifier

One of the solutions to the classification problem are the ensemble methods, in particular a hierarchical approach. This method bases on dynamically splitting the original problem during training into smaller subproblems which should be easier to train. Then the answers are combined together to obtain the final classification. The main problem here is how to divide (cluster) the original problem to obtain best possible accuracy expressed in terms of risk function value. The exact value for a given clustering is known only after the whole training process. In this paper we propose the risk estimation method based on the analysis of the root classifier. This makes it possible to evaluate the risks for all subproblems without any training of children classifiers. Together with some earlier theoretical results on hierarchical approach, we show how to use the proposed method to evaluate the risk for the whole ensemble. A variant, which uses a genetic algorithm (GA), is proposed. We compare this method with an earlier one, based on the Bayes law. We show that the subproblem risk evaluation is highly correlated with the true risk, and that the Bayes/GA approaches give hierarchical classifiers which are superior to single ones. Our method works for any classifier which returns a class probability vector for a given example.     

EVALUATION OF CLASSIFIERS FOR AN UNEVEN CLASS DISTRIBUTION PROBLEM

Classification problems with uneven class distributions present several difficulties during the training as well as during the evaluation process of classifiers. A classification problem with such characteristics has resulted from a data mining project where the objective was to predict customer insolvency. Using the data set from the customer insolvency problem, we study several alternative methodologies, which have been reported to better suit the specific characteristics of this type of problem. Three different but equally important directions are examined: (a) the performance measures that should be used for problems in this domain; (b) the class distributions that should be used for the training data sets; and (c) the classification algorithms to be used. The final evaluation of the resulting classifiers is based on a study of the economic impact of classification results. This study concludes to a framework that provides the “best” classifiers, identifies the performance measures that should be used as the decision criterion, and suggests the “best” class distribution based on the value of the relative gain from correct classification in the positive class. This framework has been applied in the customer insolvency problem, but it is claimed that it can be applied to many similar problems with uneven class distributions that almost always require a multi-objective evaluation process.     

Exploration of classification confidence in ensemble learning

Ensemble learning has attracted considerable attention owing to its good generalization performance. The main issues in constructing a powerful ensemble include training a set of diverse and accurate base classifiers, and effectively combining them. Ensemble margin, computed as the difference of the vote numbers received by the correct class and the another class received with the most votes, is widely used to explain the success of ensemble learning. This definition of the ensemble margin does not consider the classification confidence of base classifiers. In this work, we explore the influence of the classification confidence of the base classifiers in ensemble learning and obtain some interesting conclusions. First, we extend the definition of ensemble margin based on the classification confidence of the base classifiers. Then, an optimization objective is designed to compute the weights of the base classifiers by minimizing the margin induced classification loss. Several strategies are tried to utilize the classification confidences and the weights. It is observed that weighted voting based on classification confidence is better than simple voting if all the base classifiers are used. In addition, ensemble pruning can further improve the performance of a weighted voting ensemble. We also compare the proposed fusion technique with some classical algorithms. The experimental results also show the effectiveness of weighted voting with classification confidence.     

Diversity-based combination of non-parametric classifiers for EMG signal decomposition

Non-parametric classification procedures based on a certainty measure and nearest neighbour rule for motor unit potential classification (MUP) during electromyographic (EMG) signal decomposition were explored. A diversity-based classifier fusion approach is developed and evaluated to achieve improved classification performance. The developed system allows the construction of a set of non-parametric base classifiers and then automatically chooses, from the pool of base classifiers, subsets of classifiers to form candidate classifier ensembles. The system selects the classifier ensemble members by exploiting a diversity measure for selecting classifier teams. The kappa statistic is used as the diversity measure to estimate the level of agreement between base classifier outputs, i.e., to measure the degree of decision similarity between base classifiers. The pool of base classifiers consists of two kinds of classifiers: adaptive certainty-based classifiers (ACCs) and adaptive fuzzy k-NN classifiers (AFNNCs) and both utilize different types of features. Once the patterns are assigned to their classes, by the classifier fusion system, firing pattern consistency statistics for each class are calculated to detect classification errors in an adaptive fashion. Performance of the developed system was evaluated using real and simulated EMG signals and was compared with the performance of the constituent base classifiers and the performance of the fixed ensemble containing the full set of base classifiers. Across the EMG signal data sets used, the diversity-based classifier fusion approach had better average classification performance overall, especially in terms of reducing classification errors.     

Bayes-optimality motivated linear and multilayered perceptron-baseddimensionality reduction

Dimensionality reduction is the process of mapping high-dimension patterns to a lower dimension subspace. When done prior to classification, estimates obtained in the lower dimension subspace are more reliable. For some classifiers, there is also an improvement in performance due to the removal of the diluting effect of redundant information. A majority of the present approaches to dimensionality reduction are based on scatter matrices or other statistics of the data which do not directly correlate to classification accuracy. The optimality criteria of choice for the purposes of classification is the Bayes error. Usually however, Bayes error is difficult to express analytically. We propose an optimality criteria based on an approximation of the Bayes error and use it to formulate a linear and a nonlinear method of dimensionality reduction. The nonlinear method we propose, relies on using a multilayered perceptron which produces as output the lower dimensional representation. It thus differs from autoassociative like multilayered perceptrons which have been proposed and used for dimensionality reduction. Our results show that the nonlinear method is, as anticipated, superior to the linear method in that it can perform unfolding of a nonlinear manifold. In addition, the nonlinear method we propose provides substantially better lower dimension representation (for classification purposes) than Fisher's linear discriminant (FLD) and two other nonlinear methods of dimensionality reduction that are often used.     

Balanced undersampling: a novel sentence-based undersampling method to improve recognition of named entities in chemical and biomedical text

The class imbalance problem is a key factor that affects the performance of many classification tasks when using machine learning methods. This mainly refers to the problem where the number of samples in certain classes is much greater than in others. Such imbalance considerably affects the performance of classifiers in which the majority class or classes are often favored, thus resulting in high-precision/low-recall classifiers. Named entity recognition in free text suffers from this problem to a large extent because in any given free text, many samples do not belong to a specific entity. Furthermore, the data used in this specific type of classification is in sequenced mode and is different than that used in other common classification tasks such as image classification, spam detection, and text classification in which no semantic or syntactic relation exists between samples. In this study, we propose an undersampling approach for sequenced data that preserves existing correlations between sequenced samples that comprise sentences and thus improve the performance of classifiers. We call this method balanced undersampling (BUS). Considering the recent increased interest in the use of NER in the chemical and biomedical domains, the proposed method is developed and tested on four recent state-of-the-art corpora in these domains, including BioCreative IV ChemDNER, Bio-entity Recognition Challenge of JNLPBA (JNLPBA), SemEval2013 DDI DrugBank, and SemEval2013 DDI Medline datasets. The performance of the proposed method is evaluated against two other common undersampling methods: random undersampling and stop-word filtering. Our method is shown to outperform both methods with respect to F-score for all datasets used.     

Hierarchical classification method based on selective learning of slacked hierarchy for activity recognition systems

Physical activity recognition using wearable sensors has gained significant interest from researchers working in the field of ambient intelligence and human behavior analysis. The problem of multi-class classification is an important issue in the applications which naturally has more than two classes. A well-known strategy to convert a multi-class classification problem into binary sub-problems is the error-correcting output coding (ECOC) method. Since existing methods use a single classifier with ECOC without considering the dependency among multiple classifiers, it often fails to generalize the performance and parameters in a real-life application, where different numbers of devices, sensors and sampling rates are used. To address this problem, we propose a unique hierarchical classification model based on the combination of two base binary classifiers using selective learning of slacked hierarchy and integrating the training of binary classifiers into a unified objective function. Our method maps the multi-class classification problem to multi-level classification. A multi-tier voting scheme has been introduced to provide a final classification label at each level of the solicited model. The proposed method is evaluated on two publicly available datasets and compared with independent base classifiers. Furthermore, it has also been tested on real-life sensor readings for 3 different subjects to recognize four activities i.e. Walking, Standing, Jogging and Sitting. The presented method uses same hierarchical levels and parameters to achieve better performance on all three datasets having different number of devices, sensors and sampling rates. The average accuracies on publicly available dataset and real-life sensor readings were recorded to be 95% and 85%, respectively. The experimental results validate the effectiveness and generality of the proposed method in terms of performance and parameters.     

Risk-sensitive loss functions for sparse multi-category classification problems

In this paper, we propose two risk-sensitive loss functions to solve the multi-category classification problems where the number of training samples is small and/or there is a high imbalance in the number of samples per class. Such problems are common in the bio-informatics/medical diagnosis areas. The most commonly used loss functions in the literature do not perform well in these problems as they minimize only the approximation error and neglect the estimation error due to imbalance in the training set. The proposed risk-sensitive loss functions minimize both the approximation and estimation error. We present an error analysis for the risk-sensitive loss functions along with other well known loss functions. Using a neural architecture, classifiers incorporating these risk-sensitive loss functions have been developed and their performance evaluated for two real world multi-class classification problems, viz., a satellite image classification problem and a micro-array gene expression based cancer classification problem. To study the effectiveness of the proposed loss functions, we have deliberately imbalanced the training samples in the satellite image problem and compared the performance of our neural classifiers with those developed using other well-known loss functions. The results indicate the superior performance of the neural classifier using the proposed loss functions both in terms of the overall and per class classification accuracy. Performance comparisons have also been carried out on a number of benchmark problems where the data is normal i.e., not sparse or imbalanced. Results indicate similar or better performance of the proposed loss functions compared to the well-known loss functions.     

Benchmarking Classification Models for Software Defect Prediction: A Proposed Framework and Novel Findings

Software defect prediction strives to improve software quality and testing efficiency by constructing predictive classification models from code attributes to enable a timely identification of fault-prone modules. Several classification models have been evaluated for this task. However, due to inconsistent findings regarding the superiority of one classifier over another and the usefulness of metric-based classification in general, more research is needed to improve convergence across studies and further advance confidence in experimental results. We consider three potential sources for bias: comparing classifiers over one or a small number of proprietary datasets, relying on accuracy indicators that are conceptually inappropriate for software defect prediction and cross-study comparisons, and finally, limited use of statisti-cal testing procedures to secure empirical findings. To remedy these problems, a framework for comparative software defect prediction experiments is proposed and applied in a large-scale empirical comparison of 22 classifiers over ten public domain datasets from the NASA Metrics Data repository. Our results indicate that the importance of the particu-lar classification algorithm may have been overestimated in previous research since no significant performance differ-ences could be detected among the top-17 classifiers.     

ECG quality assessment based on a kernel support vector machine and genetic algorithm with a feature matrix

We propose a systematic ECG quality classification method based on a kernel support vector machine（KSVM） and genetic algorithm（GA） to determine whether ECGs collected via mobile phone are acceptable or not. This method includes mainly three modules, i.e., lead-fall detection, feature extraction, and intelligent classification. First, lead-fall detection is executed to make the initial classification. Then the power spectrum, baseline drifts, amplitude difference, and other time-domain features for ECGs are analyzed and quantified to form the feature matrix. Finally, the feature matrix is assessed using KSVM and GA to determine the ECG quality classification results. A Gaussian radial basis function（GRBF） is employed as the kernel function of KSVM and its performance is compared with that of the Mexican hat wavelet function（MHWF）. GA is used to determine the optimal parameters of the KSVM classifier and its performance is compared with that of the grid search（GS） method. The performance of the proposed method was tested on a database from PhysioNet/Computing in Cardiology Challenge 2011, which includes 1500 12-lead ECG recordings. True positive（TP）, false positive（FP）, and classification accuracy were used as the assessment indices. For training database set A（1000 recordings）, the optimal results were obtained using the combination of lead-fall, GA, and GRBF methods, and the corresponding results were： TP 92.89%, FP 5.68%, and classification accuracy 94.00%. For test database set B（500 recordings）, the optimal results were also obtained using the combination of lead-fall, GA, and GRBF methods, and the classification accuracy was 91.80%.     

Farsi/Arabic handwritten digit recognition based on ensemble of SVD classifiers and reliable multi-phase PSO combination rule

The problem of handwritten digit recognition has long been an open problem in the field of pattern classification and of great importance in industry. The heart of the problem lies within the ability to design an efficient algorithm that can recognize digits written and submitted by users via a tablet, scanner, and other digital devices. From an engineering point of view, it is desirable to achieve a good performance within limited resources. To this end, we have developed a new approach for handwritten digit recognition that uses a small number of patterns for training phase. To improve the overall performance achieved in classification task, the literature suggests combining the decision of multiple classifiers rather than using the output of the best classifier in the ensemble; so, in this new approach, an ensemble of classifiers is used for the recognition of handwritten digit. The classifiers used in proposed system are based on singular value decomposition (SVD) algorithm. The experimental results and the literature show that the SVD algorithm is suitable for solving sparse matrices such as handwritten digit. The decisions obtained by SVD classifiers are combined by a novel proposed combination rule which we named reliable multi-phase particle swarm optimization. We call the method “Reliable” because we have introduced a novel reliability parameter which is applied to tackle the problem of PSO being trapped in local minima. In comparison with previous methods, one of the significant advantages of the proposed method is that it is not sensitive to the size of training set. Unlike other methods, the proposed method uses just 15 % of the dataset as a training set, while other methods usually use (60–75) % of the whole dataset as the training set. To evaluate the proposed method, we tested our algorithm on Farsi/Arabic handwritten digit dataset. What makes the recognition of the handwritten Farsi/Arabic digits more challenging is that some of the digits can be legally written in different shapes. Therefore, 6000 hard samples (600 samples per class) are chosen by K-nearest neighbor algorithm from the HODA dataset which is a standard Farsi/Arabic digit dataset. Experimental results have shown that the proposed method is fast, accurate, and robust against the local minima of PSO. Finally, the proposed method is compared with state of the art methods and some ensemble classifier based on MLP, RBF, and ANFIS with various combination rules.