A new maximum margin algorithm for one-class problems and its boosting implementation

被引：20

作者：

Tao, Q ^{[1
]}

Wu, GW ^{[1
]}

Wang, J ^{[1
]}

机构：

[1] Chinese Acad Sci, Inst Automat, Key Lab Complex Syst & Intelligence Sci, Beijing 100080, Peoples R China

来源：

PATTERN RECOGNITION | 2005年 / 38卷 / 07期

基金：

中国国家自然科学基金;

关键词：

one-class problems; outliers; statistical learning theory; support vector machines; margin; boosting;

D O I：

10.1016/j.patcog.2004.10.010

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

In this paper, each one-class problem is regarded as trying to estimate a function that is positive on a desired slab and negative on the complement. The main advantage of this viewpoint is that the loss function and the expected risk can be defined to ensure that the slab can contain as many samples as possible. Inspired by the nature of SVMs, the intuitive margin is also defined. As a result, a new linear optimization problem to maximize the margin and some theoretically motivated learning algorithms are obtained. Moreover, the proposed algorithms can be implemented by boosting techniques to solve nonlinear one-class classifications. (c) 2005 Pattern Recognition Society. Published by Elsevier Ltd. All rights reserved.

引用

页码：1071 / 1077

页数：7

共 13 条

[1]

Cristianini N., 2000, Intelligent Data Analysis: An Introduction, DOI 10.1017/CBO9780511801389

[2] Linear programming boosting via column generation [J].

Demiriz, A ;

Bennett, KP ;

Shawe-Taylor, J .

MACHINE LEARNING, 2002, 46 (1-3) :225-254

[3] Arbitrary-norm separating plane [J].

Mangasarian, OL .

OPERATIONS RESEARCH LETTERS, 1999, 24 (1-2) :15-23

[4] Constructing boosting algorithms from SVMs:: An application to one-class classification [J].

Rätsch, G ;

Mika, S ;

Schölkopf, B ;

Müller, KR .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2002, 24 (09) :1184-1199

[5]

RATSCH G, 2001, THESIS U POSDAM

[6] Estimating the support of a high-dimensional distribution [J].

Schölkopf, B ;

Platt, JC ;

Shawe-Taylor, J ;

Smola, AJ ;

Williamson, RC .

NEURAL COMPUTATION, 2001, 13 (07) :1443-1471

[7] New support vector algorithms [J].

Schölkopf, B ;

Smola, AJ ;

Williamson, RC ;

Bartlett, PL .

NEURAL COMPUTATION, 2000, 12 (05) :1207-1245

[8] Regularized principal manifolds [J].

Smola, AJ ;

Mika, S ;

Schölkopf, B ;

Williamson, RC .

JOURNAL OF MACHINE LEARNING RESEARCH, 2001, 1 (03) :179-209

[9] Support vector domain description [J].

Tax, DMJ ;

Duin, RPW .

PATTERN RECOGNITION LETTERS, 1999, 20 (11-13) :1191-1199

[10] Support vector data description [J].

Tax, DMJ ;

Duin, RPW .

MACHINE LEARNING, 2004, 54 (01) :45-66

← 1 2 →