GENERATING FUZZY RULES FROM TRAINING DATA CONTAINING NOISE FOR HANDLING CLASSIFICATION PROBLEMS
Shyi‐Ming Chen, Cheng-Hsuan Kao, CHENG-HAO YU · Cybernetics & Systems · 2002
It is obvious that one of the important tasks in a fuzzy system is to find a set of rules to deal with a specific classification problem. In recent years, many researchers focused on the research topic of generating fuzzy rules from training data for handling classification problems. In a previous paper, we presented an algorithm to construct membership functions and to generate fuzzy rules from training examples. In this paper, we extend that work to propose a new algorithm to generate fuzzy rules from training data containing noise to deal with classification problems. The proposed algorithm gets a higher classification accuracy rate and generates fewer fuzzy rules and fewer input attributes in the antecedent portions of the generated fuzzy rules.