Learning Fuzzy If-Then Rules for Pattern Classi cation with Weighted Training Patterns

In this paper we propose a learning method of fuzzy if-then rules for pattern classification problems. We assume that each training pattern has a weight that describes its importance. The antecedent part of fuzzy if-then rules are specified by partitioning each attributes into fuzzy sets while the consequent class and the degree of certainty of the fuzzy if-then rules are determined from the compatibility and weights of training patterns. The proposed learning method adjusts the degree of certainty so that the classification cost is minimized. Experimental results on several UCI data sets show the effectiveness of the proposed method.