Uma abordagem multi-objetivo do método de fertilização in vitro para os algoritmos NSGA-II e GDE3

Detalhes bibliográficos
Ano de defesa: 2020
Autor(a) principal: Sampaio, Sávio Menezes lattes
Orientador(a): Camilo Junior, Celso Gonçalves lattes
Banca de defesa: Camilo Junior, Celso Gonçalves, Soares, Telma Woerle de Lima, Lima Neto, Fernando Buarque de
Tipo de documento: Dissertação
Tipo de acesso: Acesso aberto
Idioma: por
Instituição de defesa: Universidade Federal de Goiás
Programa de Pós-Graduação: Programa de Pós-graduação em Ciência da Computação (INF)
Departamento: Instituto de Informática - INF (RG)
País: Brasil
Palavras-chave em Português:
Palavras-chave em Inglês:
Área do conhecimento CNPq:
Link de acesso: http://repositorio.bc.ufg.br/tede/handle/tede/10373
Resumo: objective problems, especially for complex and multimodal. Due to the balance between its exploration and exploitation capabilities, and its ability to avoid local optimal, we speculate that this method can also improve Multi-Objective Evolutionary Algorithms. In this way, this work proposes the adaptation of the In Vitro Fertilization method to the Multi-Objective approach, with new collection and transfer criteria, as well as its coupling to the Multi-Objective Evolutionary Algorithms NSGA-II, based on Genetic Algorithms, and GDE3, based on Differential Evolution, to create new Multi-Objective Memetic Algorithms: IVF/NSGA-II and IVF/GDE3. We evaluated the efficacy of the proposals by comparing canonic NSGA-II with memetic IVF/NSGA-II, as well as GDE3 with memetic IVF/GDE3, applied to Multi-Objective benchmark ZDT, and to the Multi-Objective problem MOTSP-VENDOR. The results show that the In Vitro Fertilization method adapted to the Multi-Objective approach contributed to the fact that the memetic versions exceeded the canonical versions. The results also indicate that this approach is promising to support MOEAs.