支持向量机的新分类器算法研究(英文)  

Research on New Classifier Algorithms for Support Vector Machine

作　　者：宋晓宁[1] 吴小俊[2] 杨静宇[1] 郑宇杰[3] 杨习贝[1] 

机构地区：[1]南京理工大学计算机系,南京210094 [2]江南大学信息工程学院,无锡214122 [3]中国电子科技集团公司第28研究所,南京210007

出　　处：《系统仿真学报》2009年第12期3617-3621,共5页Journal of System Simulation

基　　金：The National Science Foundation of China (60632050, 60572034);The Natural Science Foundation of Jiangsu Province (BK2006081)

摘　　要：提出了基于支持向量机(SVM)和决策树(DT)的几种新分类器算法。首先对支持向量机和模糊支持向量机(FSVM)处理多类分类问题的传统算法进行介绍,然后在DT-FSVM和最近邻分类器的基础之上,提出基于样本区域分析(SRA)的混合分类器算法(SRA-DT-FSVM),该算法中样本区域分析的概念被提出。在线性鉴别分析(LDA)的人脸特征空间中,分别对算法FSVM,DT-FSVM,SRA-DT-FSVM以及Binary-DT-SVM进行了全面的性能比较。通过在ORL人脸库上的实验结果表明,提出的新方法是有效的。此外,基于DT的SVM的分类速度获得了较大提高。New approaches to face classifier based on support vector machine （SVM） and decision tree （DT） were developed. Firstly, traditional methods of SVM and fuzzy support vector machine （FSVM） for multi-class problems were introduced, then a hybrid classifier algorithm（SRA-DT-FSVM） based on DT-FSVM and nearest-neighbor classifier was proposed, in which the idea of sample region analysis（SRA） was incorporated into the design of the new classifiers. Meanwhile, in the LDA feature space, a comprehensive performance comparison of face recognition between the algorithms including FSVM, DT-FSVM, SRA-DT-FSVM and Binary-DT-SVM was made respectively. The experimental results conducted on ORL face images show the effectiveness of these new classifier methods. Furthermore, the classification speed of SVM based on DT can be improved a lot.

关 键 词：模糊支持向量机 决策树 样本区域分析 特征抽取 人脸识别 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]