“The Joint Selection of Features and Training Samples using Genetic Algorithms and Its Application to Character Recognition”

Akira Suzukit, Masashi Morimoto, Yoshimasa Kimura, Satoshi Shimada, Shunichi Yonemura

Research output: Contribution to journalArticle

Abstract

In this paper, we propose an improved feature selection method based on genetic algorithms (GA) that selects not only features but also training samples when creating reference patterns. In this method, the chromosome is composed of a set of bits that represents the state of feature selection. The reference pattern in each category is created from the selected training samples, and the discrimination rate for all training samples is used as the fitness value in GA. GA is used as a tool to find the suboptimal combination that increases the discrimination rate from the enormous number of combinations of features and training samples. We perform experiments using five sets of hand-written Kanji character patterns. Each set consists of patterns in two similar categories. The results show that the proposed method, which selects features and training samples, provides higher recognition rates than the conventional method that selects only features.

Original languageEnglish
Pages (from-to)650-659
Number of pages10
JournalJournal of the Institute of Image Electronics Engineers of Japan
Volume40
Issue number4
DOIs
Publication statusPublished - 2011
Externally publishedYes

Fingerprint

Character recognition
Genetic algorithms
Feature extraction
Chromosomes
Experiments

Keywords

  • character recognition
  • feature selection
  • genetic algorithms
  • training sample selection

ASJC Scopus subject areas

  • Computer Science (miscellaneous)
  • Electrical and Electronic Engineering

Cite this

“The Joint Selection of Features and Training Samples using Genetic Algorithms and Its Application to Character Recognition”. / Suzukit, Akira; Morimoto, Masashi; Kimura, Yoshimasa; Shimada, Satoshi; Yonemura, Shunichi.

In: Journal of the Institute of Image Electronics Engineers of Japan, Vol. 40, No. 4, 2011, p. 650-659.

Research output: Contribution to journalArticle

@article{035e5cdf5bdc4f0d8031d87199f47a61,
title = "“The Joint Selection of Features and Training Samples using Genetic Algorithms and Its Application to Character Recognition”",
abstract = "In this paper, we propose an improved feature selection method based on genetic algorithms (GA) that selects not only features but also training samples when creating reference patterns. In this method, the chromosome is composed of a set of bits that represents the state of feature selection. The reference pattern in each category is created from the selected training samples, and the discrimination rate for all training samples is used as the fitness value in GA. GA is used as a tool to find the suboptimal combination that increases the discrimination rate from the enormous number of combinations of features and training samples. We perform experiments using five sets of hand-written Kanji character patterns. Each set consists of patterns in two similar categories. The results show that the proposed method, which selects features and training samples, provides higher recognition rates than the conventional method that selects only features.",
keywords = "character recognition, feature selection, genetic algorithms, training sample selection",
author = "Akira Suzukit and Masashi Morimoto and Yoshimasa Kimura and Satoshi Shimada and Shunichi Yonemura",
year = "2011",
doi = "10.11371/iieej.40.650",
language = "English",
volume = "40",
pages = "650--659",
journal = "Journal of the Institute of Image Electronics Engineers of Japan",
issn = "0285-9831",
publisher = "Institute of Image Electronics Engineers of Japan",
number = "4",

}

TY - JOUR

T1 - “The Joint Selection of Features and Training Samples using Genetic Algorithms and Its Application to Character Recognition”

AU - Suzukit, Akira

AU - Morimoto, Masashi

AU - Kimura, Yoshimasa

AU - Shimada, Satoshi

AU - Yonemura, Shunichi

PY - 2011

Y1 - 2011

N2 - In this paper, we propose an improved feature selection method based on genetic algorithms (GA) that selects not only features but also training samples when creating reference patterns. In this method, the chromosome is composed of a set of bits that represents the state of feature selection. The reference pattern in each category is created from the selected training samples, and the discrimination rate for all training samples is used as the fitness value in GA. GA is used as a tool to find the suboptimal combination that increases the discrimination rate from the enormous number of combinations of features and training samples. We perform experiments using five sets of hand-written Kanji character patterns. Each set consists of patterns in two similar categories. The results show that the proposed method, which selects features and training samples, provides higher recognition rates than the conventional method that selects only features.

AB - In this paper, we propose an improved feature selection method based on genetic algorithms (GA) that selects not only features but also training samples when creating reference patterns. In this method, the chromosome is composed of a set of bits that represents the state of feature selection. The reference pattern in each category is created from the selected training samples, and the discrimination rate for all training samples is used as the fitness value in GA. GA is used as a tool to find the suboptimal combination that increases the discrimination rate from the enormous number of combinations of features and training samples. We perform experiments using five sets of hand-written Kanji character patterns. Each set consists of patterns in two similar categories. The results show that the proposed method, which selects features and training samples, provides higher recognition rates than the conventional method that selects only features.

KW - character recognition

KW - feature selection

KW - genetic algorithms

KW - training sample selection

UR - http://www.scopus.com/inward/record.url?scp=85024740467&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85024740467&partnerID=8YFLogxK

U2 - 10.11371/iieej.40.650

DO - 10.11371/iieej.40.650

M3 - Article

VL - 40

SP - 650

EP - 659

JO - Journal of the Institute of Image Electronics Engineers of Japan

JF - Journal of the Institute of Image Electronics Engineers of Japan

SN - 0285-9831

IS - 4

ER -