SINGLE-OBJECTIVE AND 2-OBJECTIVE GENETIC ALGORITHMS FOR SELECTING LINGUISTIC RULES FOR PATTERN-CLASSIFICATION PROBLEMS

This paper proposes various methods for constructing a compact fuzzy c lassification system consisting of a small number of linguistic classi fication rules. First we formulate a rule selection problem of linguis tic classification rules with two objectives: to maximize the number o f correctly classified training patterns and to minimize the number of selected rules. Next we propose three methods for finding a set of no n-dominated solutions of the rule selection problem. These three metho ds are based on a single-objective genetic algorithm. We also propose a method based on a multi-objective genetic algorithm for finding a se t of non-dominated solutions. We examine the performance of the propos ed methods by applying them to the well-known iris data. Finally we pr opose a hybrid algorithm by combining a learning method of linguistic classification rules with the multi-objective genetic algorithm. High performance of the hybrid algorithm is demonstrated by computer simula tions on the iris data. (C) 1997 Elsevier Science B.V.