Improvements in the identification of interpretable fuzzy models with exceptions based on ant colony optimization

In a previous work, the authors proposed, on one hand, an extension on the syntax of fuzzy rules by including new predicates and exceptional rules and, on the other hand, the use of an ant to obtain an optimal set of such rules that describes an initial fuzzy model. The present work proposes several extensions on that algorithm in order to improve the interpretability of the obtained fuzzy model, as well as the computational cost of the algorithm. Experimental results on several initial fuzzy models reveal the gain obtained with each extension and when applied altogether.