Clustering in product space for fuzzy inference

Abstract

The authors present an algorithm that generates a set of fuzzy rules with linear consequents from raw data using radial basis functions and an extended clustering approach. The algorithm uses output information in conjunction with adding and pruning neurons to generate a compact structure and its rough approximation quickly from one pass over the data. It is shown that this algorithm can approximate a typical nonlinear switching function by generating a set of fuzzy rules.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>