On the Design of an ECOC-Compliant Genetic Algorithm

Genetic Algorithms (GA) have been previously applied to Error-Correcting Output Codes (ECOC) in state-of-the-art works in order to find a suitable coding matrix. Nevertheless, none of the presented techniques directly take into account the properties of the ECOC matrix. As a result the considered se...

Descripción completa

Detalles Bibliográficos
Autores: Bautista Martín, Miguel Ángel, Escalera Guerrero, Sergio, Pujol Vila, Oriol, Baró i Solé, Xavier
Tipo de recurso: artículo
Estado:Versión aceptada para publicación
Fecha de publicación:2014
País:España
Institución:Universidad de Barcelona
Repositorio:Dipòsit Digital de la UB
OAI Identifier:oai:diposit.ub.edu:2445/119122
Acceso en línea:https://hdl.handle.net/2445/119122
Access Level:acceso abierto
Palabra clave:Algorismes genètics
Genetic algorithms
id ES_f8b2223c8fd45bd2f8dfc3b516e0f201
oai_identifier_str oai:diposit.ub.edu:2445/119122
network_acronym_str ES
network_name_str España
repository_id_str
spelling On the Design of an ECOC-Compliant Genetic AlgorithmBautista Martín, Miguel ÁngelEscalera Guerrero, SergioPujol Vila, OriolBaró i Solé, XavierAlgorismes genèticsGenetic algorithmsGenetic Algorithms (GA) have been previously applied to Error-Correcting Output Codes (ECOC) in state-of-the-art works in order to find a suitable coding matrix. Nevertheless, none of the presented techniques directly take into account the properties of the ECOC matrix. As a result the considered search space is unnecessarily large. In this paper, a novel Genetic strategy to optimize the ECOC coding step is presented. This novel strategy redefines the usual crossover and mutation operators in order to take into account the theoretical properties of the ECOC framework. Thus, it reduces the search space and lets the algorithm to converge faster. In addition, a novel operator that is able to enlarge the code in a smart way is introduced. The novel methodology is tested on several UCI datasets and four challenging computer vision problems. Furthermore, the analysis of the results done in terms of performance, code length and number of Support Vectors shows that the optimization process is able to find very efficient codes, in terms of the trade-off between classification performance and the number of classifiers. Finally, classification performance per dichotomizer results shows that the novel proposal is able to obtain similar or even better results while defining a more compact number of dichotomies and SVs compared to state-of-the-art approaches.Elsevier Ltd2014info:eu-repo/semantics/articleinfo:eu-repo/semantics/acceptedVersionapplication/pdfhttps://hdl.handle.net/2445/119122Articles publicats en revistes (Matemàtiques i Informàtica)reponame:Dipòsit Digital de la UBinstname:Universidad de BarcelonaInglésVersió postprint del document publicat a: https://doi.org/10.1016/j.patcog.2013.06.019Pattern Recognition, 2014, vol. 47, num. 2, p. 865-884https://doi.org/10.1016/j.patcog.2013.06.019(c) Elsevier Ltd, 2014info:eu-repo/semantics/openAccessoai:diposit.ub.edu:2445/1191222026-05-27T06:46:51Z
dc.title.none.fl_str_mv On the Design of an ECOC-Compliant Genetic Algorithm
title On the Design of an ECOC-Compliant Genetic Algorithm
spellingShingle On the Design of an ECOC-Compliant Genetic Algorithm
Bautista Martín, Miguel Ángel
Algorismes genètics
Genetic algorithms
title_short On the Design of an ECOC-Compliant Genetic Algorithm
title_full On the Design of an ECOC-Compliant Genetic Algorithm
title_fullStr On the Design of an ECOC-Compliant Genetic Algorithm
title_full_unstemmed On the Design of an ECOC-Compliant Genetic Algorithm
title_sort On the Design of an ECOC-Compliant Genetic Algorithm
dc.creator.none.fl_str_mv Bautista Martín, Miguel Ángel
Escalera Guerrero, Sergio
Pujol Vila, Oriol
Baró i Solé, Xavier
author Bautista Martín, Miguel Ángel
author_facet Bautista Martín, Miguel Ángel
Escalera Guerrero, Sergio
Pujol Vila, Oriol
Baró i Solé, Xavier
author_role author
author2 Escalera Guerrero, Sergio
Pujol Vila, Oriol
Baró i Solé, Xavier
author2_role author
author
author
dc.subject.none.fl_str_mv Algorismes genètics
Genetic algorithms
topic Algorismes genètics
Genetic algorithms
description Genetic Algorithms (GA) have been previously applied to Error-Correcting Output Codes (ECOC) in state-of-the-art works in order to find a suitable coding matrix. Nevertheless, none of the presented techniques directly take into account the properties of the ECOC matrix. As a result the considered search space is unnecessarily large. In this paper, a novel Genetic strategy to optimize the ECOC coding step is presented. This novel strategy redefines the usual crossover and mutation operators in order to take into account the theoretical properties of the ECOC framework. Thus, it reduces the search space and lets the algorithm to converge faster. In addition, a novel operator that is able to enlarge the code in a smart way is introduced. The novel methodology is tested on several UCI datasets and four challenging computer vision problems. Furthermore, the analysis of the results done in terms of performance, code length and number of Support Vectors shows that the optimization process is able to find very efficient codes, in terms of the trade-off between classification performance and the number of classifiers. Finally, classification performance per dichotomizer results shows that the novel proposal is able to obtain similar or even better results while defining a more compact number of dichotomies and SVs compared to state-of-the-art approaches.
publishDate 2014
dc.date.none.fl_str_mv 2014
dc.type.none.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/acceptedVersion
format article
status_str acceptedVersion
dc.identifier.none.fl_str_mv https://hdl.handle.net/2445/119122
url https://hdl.handle.net/2445/119122
dc.language.none.fl_str_mv Inglés
language_invalid_str_mv Inglés
dc.relation.none.fl_str_mv Versió postprint del document publicat a: https://doi.org/10.1016/j.patcog.2013.06.019
Pattern Recognition, 2014, vol. 47, num. 2, p. 865-884
https://doi.org/10.1016/j.patcog.2013.06.019
dc.rights.none.fl_str_mv (c) Elsevier Ltd, 2014
info:eu-repo/semantics/openAccess
rights_invalid_str_mv (c) Elsevier Ltd, 2014
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
dc.publisher.none.fl_str_mv Elsevier Ltd
publisher.none.fl_str_mv Elsevier Ltd
dc.source.none.fl_str_mv Articles publicats en revistes (Matemàtiques i Informàtica)
reponame:Dipòsit Digital de la UB
instname:Universidad de Barcelona
instname_str Universidad de Barcelona
reponame_str Dipòsit Digital de la UB
collection Dipòsit Digital de la UB
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869425025634795520
score 15,301603

Similares en LA Referencia