We present an approach to Case-Based Reasoning grounded on fuzzy relations and residuated implications operators. We propose to create clusters of cases in the base using fuzzy gradual rules, modelling the principle ``the more similar the problem descriptions are, the more similar the solution descriptions are''. We also study the use of the Fuzzy ART neural network to create the clusters. We study a set of strategies to obtain weights for cases in the training base, considering the existence of not of clusters, and how to calculate the solution to a new problem. The obtained results for a real-world application are superior than those from the literature.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados