IMPROVEMENT OF THE TWO DATA CLUSTERING ALGORITHMS: FCM & AVQ AND APPLICATION TO APPROXIMATE FUZZY SYSTEM

Abstract

Knowledge discovery is one of the most important problems in the knowledge engeneering domain. Where, data clustering algorithms take the prerequisite role to the quality of received knowledge. This paper presents some improments in the two data clustering algorithms FCM and AVQ. Then there is, in the same context, also an applying of the research result to the unsupervisor learning phase of the SAM system of correlational approximation.