On the implication of equivalence of fuzzy systems to neural networks

Although the equivalence between fuzzy and neural systems is considered in various aspects depending on the context, the real implication however, of this equivalence is not explicitly addressed. As result of this, unless one is expert on both the fuzzy logic and neural network fields, there is no clear indication what circumstances prevail to implement any of them. The aim of this paper is to address this ambivalence in the context of fuzzy modeling. By means of same regression formalism the equivalence of fuzzy systems and neural networks for data-driven modeling is investigated, and a firm understanding about the merits of utilization of each system for modeling is presented.