Genetic Fuzzy System Based on Improved Fuzzy Functions

Abstract

Fuzzy inference systems based on fuzzy rule bases (FRBs) have been successfully used to model real problems. Some of the limitations exhibited by these traditional fuzzy inference systems are that there is an abundance of fuzzy operations and operators that an expert should identify. In this paper we present an alternate learning and reasoning schema, which use fuzzy functions instead of if. then rule base structures. The new fuzzy function approach optimized with genetic algorithms is proposed to replace the fuzzy operators and operations of FRBs and improve accuracy of the fuzzy models. The structure identification of the new approach is based on a supervised hybrid fuzzy clustering, entitled Improved Fuzzy Clustering (IFC) method, which yields improved membership values. The merit of the proposed fuzzy functions method is that the uncertain information on natural grouping of data samples, i.e., membership values, is utilized as additional predictors while structuring fuzzy functions and optimized with evolutionary methods. The comparative experiments using real manufacturing and financial datasets demonstrate that the proposed method is comparable or better in modeling systems of regression problem domains.