Minimizing the fuzzy rule base and maximizing its performance by a multiobjective genetic algorithm

In this paper, we explain how a GA-based multiobjective fuzzy rule selection method can be applied to high-dimensional pattern classification problems with many continuous attributes. Our rule selection method has two objectives to minimize the number of selected fuzzy if-then rules and to maximize the number of correctly classified patterns. Since the number of candidate fuzzy if-then rules in the rule selection method exponentially increases as the number of attributes increases, the rule selection method cannot handle all the fuzzy if-then rules as candidate rules when it is applied to high-dimensional pattern classification problems with many attributes. Thus we have to restrict the number of candidate rules. For this purpose, we generate only fuzzy if-then rules with a small number of antecedent conditions as candidate rules. The ability of our rule selection method is examined by computer simulations on a real-world pattern classification problem with many continuous attributes