Derivation of Fuzzy Rules from Interval-Valued Data

Abstract

Fuzzy inference systems are widely used for classification and control. They can be designed from the training data. This paper describes a technique for deriving fuzzy classification rules from the interval-valued data. The technique based on a heuristic method of possibilistic clustering and a special method of the interval-valued data preprocessing. Basic concepts of the heuristic method of possibilistic clustering based on the allotment concept are described and the method of the intervalvalued data preprocessing is also given. The method of constructing of fuzzy rules based on clustering results is presented. An illustrative example of the method’s application to the Sato and Jain’s interval-valued data is carried out. Preliminary conclusions are formulated.