Please wait a minute...
FITEE
Front. Inform. Technol. Electron. Eng.  2011, Vol. 12 Issue (5): 362-370    DOI: 10.1631/jzus.C1000361
    
Integrating outlier filtering in large margin training
Xi-chuan Zhou*,1, Hai-bin Shen2, Jie-ping Ye3
1 College of Communication Engineering, Chongqing University, Chongqing 400044, China 2 School of Electrical Engineering, Zhejiang University, Hangzhou 310027, China 3 Department of Computer Science and Engineering, Arizona State University, Tempe 85281, USA
Integrating outlier filtering in large margin training
Xi-chuan Zhou*,1, Hai-bin Shen2, Jie-ping Ye3
1 College of Communication Engineering, Chongqing University, Chongqing 400044, China 2 School of Electrical Engineering, Zhejiang University, Hangzhou 310027, China 3 Department of Computer Science and Engineering, Arizona State University, Tempe 85281, USA
 全文: PDF(433 KB)  
摘要： Large margin classifiers such as support vector machines (SVM) have been applied successfully in various classification tasks. However, their performance may be significantly degraded in the presence of outliers. In this paper, we propose a robust SVM formulation which is shown to be less sensitive to outliers. The key idea is to employ an adaptively weighted hinge loss that explicitly incorporates outlier filtering in the SVM training, thus performing outlier filtering and classification simultaneously. The resulting robust SVM formulation is non-convex. We first relax it into a semi-definite programming which admits a global solution. To improve the efficiency, an iterative approach is developed. We have performed experiments using both synthetic and real-world data. Results show that the performance of the standard SVM degrades rapidly when more outliers are included, while the proposed robust SVM training is more stable in the presence of outliers.
关键词： Support vector machines;  Outlier filter;  Semi-definite programming;  Multi-stage relaxation    
Abstract: Large margin classifiers such as support vector machines (SVM) have been applied successfully in various classification tasks. However, their performance may be significantly degraded in the presence of outliers. In this paper, we propose a robust SVM formulation which is shown to be less sensitive to outliers. The key idea is to employ an adaptively weighted hinge loss that explicitly incorporates outlier filtering in the SVM training, thus performing outlier filtering and classification simultaneously. The resulting robust SVM formulation is non-convex. We first relax it into a semi-definite programming which admits a global solution. To improve the efficiency, an iterative approach is developed. We have performed experiments using both synthetic and real-world data. Results show that the performance of the standard SVM degrades rapidly when more outliers are included, while the proposed robust SVM training is more stable in the presence of outliers.
Key words: Support vector machines    Outlier filter    Semi-definite programming    Multi-stage relaxation
收稿日期: 2010-10-15 出版日期: 2011-05-09
CLC:  TP301  
服务　　
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章　　
Xi-chuan Zhou
Hai-bin Shen
Jie-ping Ye

引用本文:

Xi-chuan Zhou, Hai-bin Shen, Jie-ping Ye. Integrating outlier filtering in large margin training. Front. Inform. Technol. Electron. Eng., 2011, 12(5): 362-370.

链接本文:

http://www.zjujournals.com/xueshu/fitee/CN/10.1631/jzus.C1000361    或    http://www.zjujournals.com/xueshu/fitee/CN/Y2011/V12/I5/362