Contrasting termination criteria for genetic algorithms

Autores
Bermúdez, Carlos; Alfonso, Hugo; Gallard, Raúl Hector
Año de publicación
1999
Idioma
español castellano
Tipo de recurso
documento de conferencia
Estado
versión publicada
Descripción
To find a good termination criterion for genetic algorithms is a difficult and frequently ignored task. In most instances the practitioner stops the algorittm after a predefined number of generations or function evaluations. How this number is established? This stop criteria assume a user's knowledge on the characteristic of the function, which influence the length of the search. But usually it is difficult to say a priori that the total number of generations should be a detemined one. ConsequentIy this approach can involve a waste of computational resources, because the genetic algorithm could stagnate at some local or global optimum and no further improvement is achieved in that condition. This presentation discusses perfomance results on evolutionary algorithms optimizing four highly multimodal functions (Michalewicz's F1 and F2, Branin's Rcos, Griewank's). The genotypic and phenotypic approaches were implemented using the Grefenstette's bias b and the stability of mean population fitness as measures of convergence, respectively. Quality of results and speed of convergence are the main perfomance variables contrasted.
Eje: Redes y sistemas inteligentes
Red de Universidades con Carreras en Informática (RedUNCI)
Materia
Ciencias Informáticas
ARTIFICIAL INTELLIGENCE
Algorithms
genetic algorithms
contrasting termination criteria
Nivel de accesibilidad
acceso abierto
Condiciones de uso
http://creativecommons.org/licenses/by-nc-sa/2.5/ar/
Repositorio
SEDICI (UNLP)
Institución
Universidad Nacional de La Plata
OAI Identificador
oai:sedici.unlp.edu.ar:10915/22218

id SEDICI_360a0c7600b06659b887b3b1b867f90c
oai_identifier_str oai:sedici.unlp.edu.ar:10915/22218
network_acronym_str SEDICI
repository_id_str 1329
network_name_str SEDICI (UNLP)
spelling Contrasting termination criteria for genetic algorithmsBermúdez, CarlosAlfonso, HugoGallard, Raúl HectorCiencias InformáticasARTIFICIAL INTELLIGENCEAlgorithmsgenetic algorithmscontrasting termination criteriaTo find a good termination criterion for genetic algorithms is a difficult and frequently ignored task. In most instances the practitioner stops the algorittm after a predefined number of generations or function evaluations. How this number is established? This stop criteria assume a user's knowledge on the characteristic of the function, which influence the length of the search. But usually it is difficult to say a priori that the total number of generations should be a detemined one. ConsequentIy this approach can involve a waste of computational resources, because the genetic algorithm could stagnate at some local or global optimum and no further improvement is achieved in that condition. This presentation discusses perfomance results on evolutionary algorithms optimizing four highly multimodal functions (Michalewicz's F1 and F2, Branin's Rcos, Griewank's). The genotypic and phenotypic approaches were implemented using the Grefenstette's bias b and the stability of mean population fitness as measures of convergence, respectively. Quality of results and speed of convergence are the main perfomance variables contrasted.Eje: Redes y sistemas inteligentesRed de Universidades con Carreras en Informática (RedUNCI)1999-05info:eu-repo/semantics/conferenceObjectinfo:eu-repo/semantics/publishedVersionObjeto de conferenciahttp://purl.org/coar/resource_type/c_5794info:ar-repo/semantics/documentoDeConferenciaapplication/pdfhttp://sedici.unlp.edu.ar/handle/10915/22218spainfo:eu-repo/semantics/openAccesshttp://creativecommons.org/licenses/by-nc-sa/2.5/ar/Creative Commons Attribution-NonCommercial-ShareAlike 2.5 Argentina (CC BY-NC-SA 2.5)reponame:SEDICI (UNLP)instname:Universidad Nacional de La Platainstacron:UNLP2025-09-03T10:27:47Zoai:sedici.unlp.edu.ar:10915/22218Institucionalhttp://sedici.unlp.edu.ar/Universidad públicaNo correspondehttp://sedici.unlp.edu.ar/oai/snrdalira@sedici.unlp.edu.arArgentinaNo correspondeNo correspondeNo correspondeopendoar:13292025-09-03 10:27:47.688SEDICI (UNLP) - Universidad Nacional de La Platafalse
dc.title.none.fl_str_mv Contrasting termination criteria for genetic algorithms
title Contrasting termination criteria for genetic algorithms
spellingShingle Contrasting termination criteria for genetic algorithms
Bermúdez, Carlos
Ciencias Informáticas
ARTIFICIAL INTELLIGENCE
Algorithms
genetic algorithms
contrasting termination criteria
title_short Contrasting termination criteria for genetic algorithms
title_full Contrasting termination criteria for genetic algorithms
title_fullStr Contrasting termination criteria for genetic algorithms
title_full_unstemmed Contrasting termination criteria for genetic algorithms
title_sort Contrasting termination criteria for genetic algorithms
dc.creator.none.fl_str_mv Bermúdez, Carlos
Alfonso, Hugo
Gallard, Raúl Hector
author Bermúdez, Carlos
author_facet Bermúdez, Carlos
Alfonso, Hugo
Gallard, Raúl Hector
author_role author
author2 Alfonso, Hugo
Gallard, Raúl Hector
author2_role author
author
dc.subject.none.fl_str_mv Ciencias Informáticas
ARTIFICIAL INTELLIGENCE
Algorithms
genetic algorithms
contrasting termination criteria
topic Ciencias Informáticas
ARTIFICIAL INTELLIGENCE
Algorithms
genetic algorithms
contrasting termination criteria
dc.description.none.fl_txt_mv To find a good termination criterion for genetic algorithms is a difficult and frequently ignored task. In most instances the practitioner stops the algorittm after a predefined number of generations or function evaluations. How this number is established? This stop criteria assume a user's knowledge on the characteristic of the function, which influence the length of the search. But usually it is difficult to say a priori that the total number of generations should be a detemined one. ConsequentIy this approach can involve a waste of computational resources, because the genetic algorithm could stagnate at some local or global optimum and no further improvement is achieved in that condition. This presentation discusses perfomance results on evolutionary algorithms optimizing four highly multimodal functions (Michalewicz's F1 and F2, Branin's Rcos, Griewank's). The genotypic and phenotypic approaches were implemented using the Grefenstette's bias b and the stability of mean population fitness as measures of convergence, respectively. Quality of results and speed of convergence are the main perfomance variables contrasted.
Eje: Redes y sistemas inteligentes
Red de Universidades con Carreras en Informática (RedUNCI)
description To find a good termination criterion for genetic algorithms is a difficult and frequently ignored task. In most instances the practitioner stops the algorittm after a predefined number of generations or function evaluations. How this number is established? This stop criteria assume a user's knowledge on the characteristic of the function, which influence the length of the search. But usually it is difficult to say a priori that the total number of generations should be a detemined one. ConsequentIy this approach can involve a waste of computational resources, because the genetic algorithm could stagnate at some local or global optimum and no further improvement is achieved in that condition. This presentation discusses perfomance results on evolutionary algorithms optimizing four highly multimodal functions (Michalewicz's F1 and F2, Branin's Rcos, Griewank's). The genotypic and phenotypic approaches were implemented using the Grefenstette's bias b and the stability of mean population fitness as measures of convergence, respectively. Quality of results and speed of convergence are the main perfomance variables contrasted.
publishDate 1999
dc.date.none.fl_str_mv 1999-05
dc.type.none.fl_str_mv info:eu-repo/semantics/conferenceObject
info:eu-repo/semantics/publishedVersion
Objeto de conferencia
http://purl.org/coar/resource_type/c_5794
info:ar-repo/semantics/documentoDeConferencia
format conferenceObject
status_str publishedVersion
dc.identifier.none.fl_str_mv http://sedici.unlp.edu.ar/handle/10915/22218
url http://sedici.unlp.edu.ar/handle/10915/22218
dc.language.none.fl_str_mv spa
language spa
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
http://creativecommons.org/licenses/by-nc-sa/2.5/ar/
Creative Commons Attribution-NonCommercial-ShareAlike 2.5 Argentina (CC BY-NC-SA 2.5)
eu_rights_str_mv openAccess
rights_invalid_str_mv http://creativecommons.org/licenses/by-nc-sa/2.5/ar/
Creative Commons Attribution-NonCommercial-ShareAlike 2.5 Argentina (CC BY-NC-SA 2.5)
dc.format.none.fl_str_mv application/pdf
dc.source.none.fl_str_mv reponame:SEDICI (UNLP)
instname:Universidad Nacional de La Plata
instacron:UNLP
reponame_str SEDICI (UNLP)
collection SEDICI (UNLP)
instname_str Universidad Nacional de La Plata
instacron_str UNLP
institution UNLP
repository.name.fl_str_mv SEDICI (UNLP) - Universidad Nacional de La Plata
repository.mail.fl_str_mv alira@sedici.unlp.edu.ar
_version_ 1842260115999162368
score 13.13397