BDBComp
Parceria:
SBC
Um Estudo Comparativo de Extração de Conhecimento Simbólico de Redes Neurais

Claudia Regina MilaréAndré Carlos Ponce de Leon Ferreira de Carvalho

­ Although Artificial Neural Networks (ANNs) have been satisfactorily employed in several problems, they still suffer of significant limitations. One of them is that the induced concept representation is not usually comprehensible to humans. Several techniques have been suggested to extract meaningful knowledge from ANNs. This paper proposes the use of symbolic learning algorithms to extract symbolic representations from ANNs. The procedure proposed is similar to that used by the Trepan algorithm, which extracts comprehensible, symbolic representations (decision trees) from ANNs. Some experiments were carried to evaluate the procedure proposed. Clique no link abaixo para buscar o texto completo deste trabalho na Web: Buscar na Web

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