A feature selection method based on minimizing generalization bounds of SVM via GA

Author

Xu, J.Q. ; Yuan, Z.D.

Author_Institution

Center of Math. & Phys. Teaching, Shanghai Inst. of Technol., China

fYear

2003

fDate

8-8 Oct. 2003

Firstpage

996

Lastpage

999

Abstract

A method based on finding those features which minimizing a kind of generalization bounds of the SVM is presented. The searching can be efficiently implemented via genetic algorithm. The resulting algorithm is shown to be effective on both simulation datasets and the real cardiac pattern recognition.

Keywords

cardiology; feature extraction; genetic algorithms; support vector machines; GA; SVM; feature selection method; generalization bounds; genetic algorithm; minimization; real cardiac pattern recognition; simulation datasets; support vector machines;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligent Control. 2003 IEEE International Symposium on

Conference_Location

Houston, TX, USA

ISSN

2158-9860

Print_ISBN

0-7803-7891-1

Type

conf

DOI

10.1109/ISIC.2003.1254773

Filename

1254773

Link To Document

https://search.isc.ac/dl/search/defaultta.aspx?DTC=49&DC=2421465