Fuzzy Clustering Algorithms with Applications to Rule Extraction
Robert Babuška · Studies in fuzziness and soft computing · 2000
An overview of fuzzy clustering is given. The chapter starts with the definition of the basic notions of clustering and with a brief review of different approaches. Then, the focus is on fuzzy clustering based on the minimization of an objective function of the c-means type. Different algorithms are presented, including the Gustafson-Kessel algorithm, maximum-likelihood clustering, fuzzy c-varieties, c-regression models and possibilistic c-means. The choice of the different user-defined parameters is discussed and illustratives examples are given. Finally, the use of fuzzy clustering for rule extraction is addressed.