A connectionist approach to reorganization of knowledge structures in fuzzy systems

This paper addresses the problem of reorganizing knowledge structures in fuzzy systems. Reviewing the most accepted approaches to this problem, it is shown that capabilities such as adaptability and determination of linguistic labels and Membership Functions (MFs) are problems which are still in the offing. Within our framework, we propose an approach to modeling issues in fuzzy systems which is extendible to more complex knowledge structures compared to simple rule bases.