Weighted Multi-Output Type TS Fuzzy Reasoning Method and Its Application to Fuzzy Discriminant Analysis for Diabetes Diagnostic Data

Abstract

As an extension of TS (Takagi-Sugeno) fuzzy reasoning method, "weighted TS fuzzy reasoning method" is proposed in which the consequent part of the fuzzy rule consists of output function with weighting function, and its properties are discussed with the well-known simplified fuzzy reasoning method and TS fuzzy reasoning method. This paper also shows learning algorithms for the weighted TS fuzzy reasoning method with multiple-outputs by using the steepest decent method, and compares learning results by the simplified reasoning method and TS fuzzy reasoning method by applying to diabetes diagnostic data, and the proposed method obtains the best diagnostic results.