A dynamic overproduce-and-choose strategy for the selection of classifier ensembles

The overproduce-and-choose strategy, which is divided into the overproduction and selection phases, has traditionally focused on finding the most accurate subset of classifiers at the selection phase, and using it to predict the class of all the samples in the test data set. It is therefore, a static classifier ensemble selection strategy. In this paper, we propose a dynamic overproduce-and-choose strategy which combines optimization and dynamic selection in a two-level selection phase to allow the selection of the most confident subset of classifiers to label each test sample individually. The optimization level is intended to generate a population of highly accurate candidate classifier ensembles, while the dynamic selection level applies measures of confidence to reveal the candidate ensemble with the highest degree of confidence in the current decision. Experimental results conducted to compare the proposed method to a static overproduce-and-choose strategy and a classical dynamic classifier selection approach demonstrate that our method outperforms both these selection-based methods, and is also more efficient in terms of performance than combining the decisions of all classifiers in the initial pool.     

基于改进动态集成选择算法的乳腺肿块辅助诊断模型

动态集成选择算法中,待测样本的能力区域由固定样本组成,这会影响分类器选择,因此提出一种基于动态能力区域策略的DES-DCR-CIER算法。首先采用异构分类器生成基分类器池,解决同构集成分类器差异性较小和异构集成分类器数目较少的问题;然后采用相互自适应K近邻算法、逼近样本集距离中心和剔除类别边缘样本三个步骤得到待测样本的动态能力区域,基于整体互补性指数选择一组互补性强的分类器;最后通过ER规则对分类器组进行合成。在安徽合肥某三甲医院的八位超声科医生乳腺肿块诊断数据集和美国威斯康辛州乳腺癌诊断公开数据集上的实验表明,基于DES-DCR-CIER算法的诊断模型精度更优。     

Prototype selection for dynamic classifier and ensemble selection

In dynamic ensemble selection (DES) techniques, only the most competent classifiers, for the classification of a specific test sample, are selected to predict the sample’s class labels. The key in DES techniques is estimating the competence of the base classifiers for the classification of each specific test sample. The classifiers’ competence is usually estimated according to a given criterion, which is computed over the neighborhood of the test sample defined on the validation data, called the region of competence. A problem arises when there is a high degree of noise in the validation data, causing the samples belonging to the region of competence to not represent the query sample. In such cases, the dynamic selection technique might select the base classifier that overfitted the local region rather than the one with the best generalization performance. In this paper, we propose two modifications in order to improve the generalization performance of any DES technique. First, a prototype selection technique is applied over the validation data to reduce the amount of overlap between the classes, producing smoother decision borders. During generalization, a local adaptive K-Nearest Neighbor algorithm is used to minimize the influence of noisy samples in the region of competence. Thus, DES techniques can better estimate the classifiers’ competence. Experiments are conducted using 10 state-of-the-art DES techniques over 30 classification problems. The results demonstrate that the proposed scheme significantly improves the classification accuracy of dynamic selection techniques.

     

Improving dynamic facial expression recognition with feature subset selection

This paper addresses the dynamic recognition of basic facial expressions in videos using feature subset selection. Feature selection has been already used by some static classifiers where the facial expression is recognized from one single image. Past work on dynamic facial expression recognition has emphasized the issues of feature extraction and classification, however, less attention has been given to the critical issue of feature selection in the dynamic scenario. The main contributions of the paper are as follows. First, we show that dynamic facial expression recognition can be casted into a classical classification problem. Second, we combine a facial dynamics extractor algorithm with a feature selection scheme for generic classifiers.We show that the paradigm of feature subset selection with a wrapper technique can improve the dynamic recognition of facial expressions. We provide evaluations of performance on real video sequences using five standard machine learning approaches: Support Vector Machines, K Nearest Neighbor, Naive Bayes, Bayesian Networks, and Classification Trees.     

Ensemble feature selection with the simple Bayesian classification

A popular method for creating an accurate classifier from a set of training data is to build several classifiers, and then to combine their predictions. The ensembles of simple Bayesian classifiers have traditionally not been a focus of research. One way to generate an ensemble of accurate and diverse simple Bayesian classifiers is to use different feature subsets generated with the random subspace method. In this case, the ensemble consists of multiple classifiers constructed by randomly selecting feature subsets, that is, classifiers constructed in randomly chosen subspaces. In this paper, we present an algorithm for building ensembles of simple Bayesian classifiers in random subspaces. The EFS_SBC algorithm includes a hill-climbing-based refinement cycle, which tries to improve the accuracy and diversity of the base classifiers built on random feature subsets. We conduct a number of experiments on a collection of 21 real-world and synthetic data sets, comparing the EFS_SBC ensembles with the single simple Bayes, and with the boosted simple Bayes. In many cases the EFS_SBC ensembles have higher accuracy than the single simple Bayesian classifier, and than the boosted Bayesian ensemble. We find that the ensembles produced focusing on diversity have lower generalization error, and that the degree of importance of diversity in building the ensembles is different for different data sets. We propose several methods for the integration of simple Bayesian classifiers in the ensembles. In a number of cases the techniques for dynamic integration of classifiers have significantly better classification accuracy than their simple static analogues. We suggest that a reason for that is that the dynamic integration better utilizes the ensemble coverage than the static integration.     

A measure of competence based on random classification for dynamic ensemble selection

In this paper, a measure of competence based on random classification (MCR) for classifier ensembles is presented. The measure selects dynamically (i.e. for each test example) a subset of classifiers from the ensemble that perform better than a random classifier. Therefore, weak (incompetent) classifiers that would adversely affect the performance of a classification system are eliminated. When all classifiers in the ensemble are evaluated as incompetent, the classification accuracy of the system can be increased by using the random classifier instead. Theoretical justification for using the measure with the majority voting rule is given. Two MCR based systems were developed and their performance was compared against six multiple classifier systems using data sets taken from the UCI Machine Learning Repository and Ludmila Kuncheva Collection. The systems developed had typically the highest classification accuracies regardless of the ensemble type used (homogeneous or heterogeneous).     

Classifier subset selection for biomedical named entity recognition

Classifier ensembling approach is considered for biomedical named entity recognition task. A vote-based classifier selection scheme having an intermediate level of search complexity between static classifier selection and real-valued and class-dependent weighting approaches is developed. Assuming that the reliability of the predictions of each classifier differs among classes, the proposed approach is based on selection of the classifiers by taking into account their individual votes. A wide set of classifiers, each based on a different set of features and modeling parameter setting are generated for this purpose. A genetic algorithm is developed so as to label the predictions of these classifiers as reliable or not. During testing, the votes that are labeled as being reliable are combined using weighted majority voting. The classifier ensemble formed by the proposed scheme surpasses the full object F-score of the best individual classifier by 2.75% and it is the highest score achieved on the data set considered.     

Plurality voting-based multiple classifier systems: statistically independent with respect to dependent classifier sets

The simultaneous use of multiple classifiers has been shown to provide performance improvement in classification problems. The selection of an optimal set of classifiers is an important part of multiple classifier systems and the independence of classifier outputs is generally considered to be an advantage for obtaining better multiple classifier systems. In this paper, the need for the classifier independence is interrogated from classification performance point of view. The performance achieved with the use of classifiers having independent joint distributions is compared to some other classifiers which are defined to have best and worst joint distributions. These distributions are obtained by formulating the combination operation as an optimization problem. The analysis revealed several important observations about classifier selection which are then used to analyze the problem of selecting an additional classifier to be used with the available multiple classifier system.     

根据混合选择策略的直觉模糊核匹配追踪集成算法

为了从分类器集成系统中选择出一组差异性大的子分类器,从而提高集成系统的泛化能力,提出了一种基于混合选择策略的直觉模糊核匹配追踪算法.基本思想是通过扰动训练集和特征空间生成一组子分类器;然后采用k均值聚类算法将对所得子分类器进行修剪,删去其中的冗余分类器;最后根据实际识别目标动态选择出较高识别率的分类器组合,使选择性集成规模能够随识别目标的复杂程度而自适应地变化,并基于预期识别精度实现循环集成.实验结果表明,与其他常用的分类器选择方法相比,本文方法灵活高效,具有更好的识别效果和泛化能力.     

A performance comparison of mobility anchor point selection schemes in Hierarchical Mobile IPv6 networks

Hierarchical Mobile IPv6 (HMIPv6) introduces a mobility anchor point (MAP) that localizes the signaling traffic and hence reduces the handoff latency. In addition to processing binding update messages from mobile nodes (MNs) on behalf of MNs’ home agents (HAs), the MAP performs data traffic tunneling destined to or originated from MNs, both of which will burden the MAP substantially as the network size grows. To provide scalable and robust mobile Internet services to a large number of visiting MNs, multiple MAPs will be deployed. In such an environment, how to select an appropriate MAP has a vital effect on the overall network performance. In this paper, we choose four MAP selection schemes: the furthest MAP selection scheme, the nearest MAP selection scheme, the mobility-based MAP selection scheme, and the adaptive MAP selection scheme. Then, we compare their performances quantitatively in terms of signaling overhead and load balancing. It can be shown that the dynamic schemes (i.e., the mobility-based and the adaptive MAP selection schemes) are better than the static schemes (i.e., the furthest and the nearest MAP selection schemes), since the dynamic schemes can select the serving MAP depending on the MN’s characteristics, e.g., mobility and session activity. In addition, the adaptive MAP selection scheme achieves low implementation overhead and better load balancing compared with the mobility-based MAP selection scheme.     

A selection framework of sensor combination feature subset for human motion phase segmentation

Motion phase plays an important role in the spatial–temporal parameters of human motion analysis. Multi-sensor fusion technology based on inertial sensors frees the monitoring of the human body phase from space constraints and improves the flexibility of the system. However, human phase segmentation methods usually rely on the determination of the positioning of the sensor and the number of sensors, it is difficult to artificially select the number and position of the sensors, especially when human motion phases are diverse. This paper proposes a selection framework for the sensor combination feature subset for motion phase segmentation, which combines feature selection algorithms with the subsequent classifiers, and determine the optimum combination of the sensor and the feature subset according to the performance of the trained model. Through the constraint and the sensor combination feature subset (SCFS), the filter method can select any number of sensors and control the size of the feature subset; the embedded method can select any number of sensors, but the size of the feature subset is determined by the classifier model. Experimental results show that the proposed framework can effectively select a specified number of sensors without human intervention, and the number of sensors has an impact on the recognition rate of the classifier within 1.5%. In addition, the filter method has good adaptability to a variety of classifiers, and the classifier prediction time can be controlled by setting the subset size of the feature; the embedded method can achieve a better phase segmentation effect than the filter method. For the application of motion phase segmentation, the proposed framework can reliably and quickly identify redundant sensors that provide effective support for reducing the complexity of the wearable sensor system and improving user comfort.     

Ensemble component selection for improving ICA based microarray data prediction models

Independent component analysis (ICA) has been widely used to tackle the microarray dataset classification problem, but there still exists an unsolved problem that the independent component (IC) sets may not be reproducible after different ICA transformations. Inspired by the idea of ensemble feature selection, we design an ICA based ensemble learning system to fully utilize the difference among different IC sets. In this system, some IC sets are generated by different ICA transformations firstly. A multi-objective genetic algorithm (MOGA) is designed to select different biologically significant IC subsets from these IC sets, which are then applied to build base classifiers. Three schemes are used to fuse these base classifiers. The first fusion scheme is to combine all individuals in the final generation of the MOGA. In addition, in the evolution, we design a global-recording technique to record the best IC subsets of each IC set in a global-recording list. Then the IC subsets in the list are deployed to build base classifier so as to implement the second fusion scheme. Furthermore, by pruning about half of less accurate base classifiers obtained by the second scheme, a compact and more accurate ensemble system is built, which is regarded as the third fusion scheme. Three microarray datasets are used to test the ensemble systems, and the corresponding results demonstrate that these ensemble schemes can further improve the performance of the ICA based classification model, and the third fusion scheme leads to the most accurate ensemble system with the smallest ensemble size.     

一种新的启发式分类器选择方法

分类器选择是一种设计多分类器系统的有效方法,从给定候选分类器集中挑选出一个子集,使得该子集集成性能最佳。现有的分类器选择方法大多采用基于集成精度的随机搜索方法,但巨大的搜索复杂度限制了它们在更大系统中的应用。该文提出一种新的选择标准——IWCECR及一种基于IWCECR的启发式搜索算法,在手写体数字识别的实验中,从20个候选分类器中挑选子集,结果表明,该方法具有较高的搜索效率,在子集集成性能方面仅次于穷举法。     

基于最小代价的多分类器动态集成

本文提出一种基于最小代价准则的分类器动态集成方法．与一般方法不同,动态集成是根据“性能预测特征”,动态地为每一样本选择最适合的一组分类器进行集成．该选择基于使误识代价与时间代价最小化的准则,改变代价函数的定义可以方便地达到识别率与识别速度之间的不同折衷．本文中提出了两种分类器动态集成的方法,并介绍了在联机手写汉字识别中的具体应用．在实验中使了３个分类器进行动态集成,因此,得到７种分类组合．在预先定义的代价意义下,我们比较了动态集成方法和其它７种固定方法的性能．实验结果证明了动态集成方法的高灵活性、实用性和提高系统综合性能的能力．     

Local Classifier Weighting by Quadratic Programming

It has been widely accepted that the classification accuracy can be improved by combining outputs of multiple classifiers. However, how to combine multiple classifiers with various (potentially conflicting) decisions is still an open problem. A rich collection of classifier combination procedures-many of which are heuristic in nature-have been developed for this goal. In this brief, we describe a dynamic approach to combine classifiers that have expertise in different regions of the input space. To this end, we use local classifier accuracy estimates to weight classifier outputs. Specifically, we estimate local recognition accuracies of classifiers near a query sample by utilizing its nearest neighbors, and then use these estimates to find the best weights of classifiers to label the query. The problem is formulated as a convex quadratic optimization problem, which returns optimal nonnegative classifier weights with respect to the chosen objective function, and the weights ensure that locally most accurate classifiers are weighted more heavily for labeling the query sample. Experimental results on several data sets indicate that the proposed weighting scheme outperforms other popular classifier combination schemes, particularly on problems with complex decision boundaries. Hence, the results indicate that local classification-accuracy-based combination techniques are well suited for decision making when the classifiers are trained by focusing on different regions of the input space.     

Switching between selection and fusion in combining classifiers: anexperiment

This paper presents a combination of classifier selection and fusion by using statistical inference to switch between the two. Selection is applied in those regions of the feature space where one classifier strongly dominates the others from the pool [called clustering-and-selection or (CS)] and fusion is applied in the remaining regions. Decision templates (DT) method is adopted for the classifier fusion part. The proposed combination scheme (called CS+DT) is compared experimentally against its two components, and also against majority vote, naive Bayes, two joint-distribution methods (BKS and a variant due to Wernecke (1988)), the dynamic classifier selection (DCS) algorithm DCS_LA based on local accuracy (Woods et al. (1997)), and simple fusion methods such as maximum, minimum, average, and product. Based on the results with five data sets with homogeneous ensembles [multilayer perceptrons (NLPs)] and ensembles of different classifiers, we offer a discussion on when to combine classifiers and how classifier selection (static or dynamic) can be misled by the differences in the classifier team.     

Feature selection with dynamic mutual information

Feature selection plays an important role in data mining and pattern recognition, especially for large scale data. During past years, various metrics have been proposed to measure the relevance between different features. Since mutual information is nonlinear and can effectively represent the dependencies of features, it is one of widely used measurements in feature selection. Just owing to these, many promising feature selection algorithms based on mutual information with different parameters have been developed. In this paper, at first a general criterion function about mutual information in feature selector is introduced, which can bring most information measurements in previous algorithms together. In traditional selectors, mutual information is estimated on the whole sampling space. This, however, cannot exactly represent the relevance among features. To cope with this problem, the second purpose of this paper is to propose a new feature selection algorithm based on dynamic mutual information, which is only estimated on unlabeled instances. To verify the effectiveness of our method, several experiments are carried out on sixteen UCI datasets using four typical classifiers. The experimental results indicate that our algorithm achieved better results than other methods in most cases.     

一种维吾尔语联机手写识别系统

该文介绍一种维吾尔语联机手写体识别系统。其针对维吾尔语词语的书写特点采用了基于多分类器融合的系统和方法,分别使用混合高斯模型模拟整词的静态特征和隐马尔科夫模型模拟书写笔迹的动态特征,有效地提升了识别系统的准确率。在第一期实验中,整词识别率达到97%;第二期的实验中,整词识别率达到99%。     

基于分类器联合的联机图形识别方法研究

研究基于不同模式特征的多分类器联合问题，提出一种新的分类器联合方法，将该方法应用于联机几何图形的识别，实验中联合了3种分类器，每种分类器的机制各不相同，并且都基于不同的模式特征，将该分类器联合方法与现有的几种联合方法进行实验比较，实验结果表明该方法具有较高的识别率。     

Similarity learning for texture image retrieval using multiple classifier system

Multiple Classifier System has found its applications in many areas such as handwriting recognition, speaker recognition, medical diagnosis, fingerprint recognition, personal identification and others. However, there have been rare attempts to develop content-based image retrieval (CBIR) system that uses multiple classifiers to learn visual similarity. Texture as a primitive visual content is often used in many important applications (viz. Medical image analysis and medical CBIR system). In this paper, a texture image retrieval system is developed that learns the visual similarity in terms of class membership using multiple classifiers. The way proposed approach combines the decisions of multiple classifiers to obtain final class memberships of query for each of the output classes is also a novel concept. A modified distance that is weighted with the membership values obtained through similarity learning is used for ranking. Three different algorithms are proposed for the retrieval of images against a query image displaying the strength of multiple classifier approach, class membership score and their interplay to achieve the objective defined in terms of simplicity, retrieval effectiveness and speed. The proposed methods based on multiple classifiers achieve higher retrieval accuracy with lower standard deviation compared to all the competing methods irrespective of the texture database and feature set used. The multiple classifier retrieval schemes proposed here is tested for texture image retrieval. However, these can be used for any other challenging retrieval problems.