Feature subset selection using genetic algorithms for handwritten digit recognition

Author

Oliveira, L.S. ; Benahmed, N. ; Sabourin, R. ; Bortolozzi, F. ; Suen, C.Y.

Author_Institution

Laboratorio de Analise e Reconhecimento de Documentos, Pontificia Univ. Catolica do Parana, Curitiba, Brazil

fYear

2001

fDate

37165

Firstpage

362

Lastpage

369

Abstract

Two approaches using genetic algorithms for feature subset selection are compared. The first approach considers a simple genetic algorithm (SGA) while the second one takes into account an iterative genetic algorithm (IGA) which is claimed to converge faster than SGA. Initially, we present an overview of the system to be optimized and the methodology applied in the experiments as well. Next, we discuss the advantages and drawbacks of each approach based on experiments carried out on NIST SD19. Finally, we conclude that the IGA converges faster than the SGA, however, the SGA seems more suitable for our problem

Keywords

genetic algorithms; handwritten character recognition; iterative methods; optical character recognition; IGA; NIST SD19; SGA; convergence; feature subset selection; handwritten digit recognition; iterative genetic algorithm; simple genetic algorithm; Filters; Genetic algorithms; Handwriting recognition; Iterative methods; Large-scale systems; Machine intelligence; NIST; Optimization methods; Pattern recognition; Spatial databases;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer Graphics and Image Processing, 2001 Proceedings of XIV Brazilian Symposium on

Conference_Location

Florianopolis

Print_ISBN

0-7695-1330-1

Type

conf

DOI

10.1109/SIBGRAPI.2001.963077

Filename

963077