Extending CBR-ANN hybrid models using fuzzy sets

This paper presents a hybrid model to develop case-based systems, where case-based reasoning (CBR) and artificial neural networks (ANN) are now combined with fuzzy sets. The associative ANN uses fuzzy sets to process continuous attributes as linguistic variables. The case-based module justifies the problem solved by ANN using a similarity function, which includes the weights of ANN and the membership degree to defined fuzzy sets. The use of fuzzy sets enables extending the traditional crisp set, using natural language in which many words have ambiguous meanings. Experimental results show the improvement achieved using the new model.