Estudo de relações de similaridade difusas e sistemas neuro-fuzzy aplicados ao raciocínio baseado em casos

Jonas Henrique MendonçaSandra SandriFlávia Martin-BedêRicardo GuimarãesOmar Carvalho

This work is a study of an approach of case-based reasoning using fuzzy similarity relations weighted. We used several metrics to obtain the weight vectors that would later be used in data classification tasks. The main objective of this work is to show the pros and cons of relationships used. We use as a case study a problem from the real world to estimate the prevalence of schistosomiasis in a region of Brazil. The results for this test are based, in most cases, superior to approaches previously applied.

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

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: