BDBComp
Parceria:
SBC
Algoritmos Genéticos Multi-objetivo para a Seleção de Atributos

Newton SpolaôrAna Carolina LorenaHuei Diana Lee

The occurrence of irrelevant and/or redundant features in Databases can degrade the performance of computational processes for knowledge extrac- tion, motivating the application of a Feature Selection process. Multi-objective Genetic Algorithms can help identifying subsets of features which optimize com- binations of possibly conflicting feature importance measures. This paper pre- sents the use of Multi-objective Genetic Algorithms in Feature Selection, inves- tigating the use of different combinations of feature importance criteria in both labeled and unlabeled datasets.

http://www.lbd.dcc.ufmg.br/colecoes/enia/2011/0060.pdf

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato: bdbcomp@lbd.dcc.ufmg.br
     Mantida por:
LBD