Critical support vector machine without kernel function

Author

Raicharoen, T. ; Lursinsap, C.

Author_Institution

Dept. of Math., Chulalongkorn Univ., Bangkok, Thailand

Volume

5

fYear

2002

fDate

18-22 Nov. 2002

Firstpage

2532

Abstract

The drawback of SVM technique leads to a positive semidefinite quadratic programming problem with a dense, structured, positive semi-definite matrix, and also requires a set of kernel functions. We propose the learning algorithms that do not need any kernel functions. The separability is based on the critical Support vectors (CSV) essential to determine the locations of all separating hyperplanes. The algorithms give better performance compared with the other proposed SVM-based algorithms when they are tested with 2 spirals problem, Sonar, Ionosphere, Mushroom, Liver disorder, Cleveland heart, Pima diabetes, Tic Tac Toe, and Votes.

Keywords

generalisation (artificial intelligence); iterative methods; learning (artificial intelligence); minimisation; pattern classification; quadratic programming; support vector machines; critical support vector machine; generalization ability; iterative scheme; learning algorithms; optimum decision region; quadratic programming; separability; separating hyperplanes; structural risk minimization; two-spirals problem; Diabetes; Heart; Ionosphere; Kernel; Liver; Quadratic programming; Sonar; Spirals; Support vector machines; Testing;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Information Processing, 2002. ICONIP '02. Proceedings of the 9th International Conference on

Print_ISBN

981-04-7524-1

Type

conf

DOI

10.1109/ICONIP.2002.1201951

Filename

1201951