Efficient fuzzy modeling under multiple criteria by using genetic algorithm
Toshihiro Suzuki, Takeshi Furuhashi, S. Matsushita, Hiroaki Tsutsui · 2003
Fuzzy modeling is a method to describe input-output relationships of nonlinear systems. A genetic algorithm (GA) has been applied to fuzzy modeling for identification of the structure of a fuzzy model and selection of input variables. Trade-offs among multiple criteria make the search problem more complicated. For easy determination of weights on the criteria, a framework of model generation and testing was proposed by the authors. This framework divides the process of fuzzy modeling into two blocks, i.e. a model generation block and model testing block. The model generation block has criteria, with a higher degree of importance, and the model testing block has those with a lower degree of importance. In this paper, the idea of Pareto optimality is introduced to this framework and the effectiveness of the framework is examined by simulations.