SELECTION OF FUZZY IF-THEN RULES BY A GENETIC METHOD

This paper proposes a genetic algorithm-based method for constructing a fuzzy classification system with fuzzy if-then rules. In the propose d method, a rule selection problem for constructing a compact fuzzy sy stem with high classification power is formulated as a combinatorial o ptimization problem with two objectives: to maximize the classificatio n rate and to minimize the number of rules. Then a method of implement ing genetic algorithms is proposed for the application to this problem and its effectiveness is demonstrated by computer simulations. The im plementation of genetic algorithms in this paper employs an approach w here a set of fuzzy if-then rules is coded as a single individual.