Fuzzy regression analysis by a fuzzy neural network and its application to dual response optimization

Abstract

Fuzzy regression analysis achieved by a fuzzy radial basis function neural network is discussed in this paper. A fuzzy regression model constructed in such a manner is then applied to a dual-response optimization problem. Fuzzy regression models are ideally suited for dual-response optimization with two advantages: (1) many systems encountered in practice are fuzzy, and (2) fuzzy regression models have dual responses in nature. The dual response optimization problem is formulated as a multiple-objective decision-making program, and an algorithm based on the duality theory is developed to solve this problem. A numerical example is also provided for illustration.