Fuzzy Reasoning Model Using Counter Propagation and Regularization

Abstract

In this paper, we propose a fuzzy reasoning method using counter propagation network as the preprocessing and a regularization term to the energy function. Applying CPN in the preprocessing may help restrain increase the number of rules in high order problems. Furthermore, we consider that applying a regularization term is effective on the problem of local minima. The proposed method is applied to wine classification problem, and its effectiveness is discussed.