We present a hyperellipsoidal clustering method that becomes the focal point of the fuzzy modeling procedure. The aim of developing a clustering algorithm is to control the shapes of clusters flexibly. This is achieved to a great extent by introducing design and tuning parameters. We propose a simple clustering algorithm which combines hierarchical and non-hierarchical procedures, and dose not require a priori assumption on the number, centroids, and volumes of clusters.
KeywordsMembership Function Fuzzy Model Fuzzy Cluster Conditional Variable Initial Cluster
Unable to display preview. Download preview PDF.
- T. Takagi and M. Sugeno: Fuzzy Identification of Systems and Its Applications to Modeling and Control. IEEE Trans, on Systems, Man and Cybernetics, Vol.SMC-15, No.1, pp.116–132, 1985.Google Scholar
- D.W. Hosmer, Jr.: Maximum Likelihood Estimates of the Parameters of a Mixture of Two Regression Lines. Communications in Statistics, Vol.3, No. 10, pp.995–1005, 1974.Google Scholar
- R.W. Gunderson and R. Canfield: Piece-Wise Multilinear Prediction from FCV Disjoint Principal Component Models. Proc. of 3rd IFSA Congress, pp.540–543, Washington, August 6–11, 1989.Google Scholar
- Y. Tanaka and K. Wakimoto: Methods of Multivariate Statistical Analysis. Gendai-Sugakusya, (in Japanese), 1983.Google Scholar
- M.J. Box, D. Davies and W.H. Swann: Non-Linear Optimization Techniques. Obiver & Boyd, pp.52–54, 1969.Google Scholar
- D.E. Gustafson and W.C. Kessel: Fuzzy Clustering with A Fuzzy Covariance Matrix. Proc. IEEE CDC, pp.761–766, San Diago, CA, 1979.Google Scholar
- R.N. Dave: An Adaptive Fuzzy c-Elliptotype Clustering Algorithm. Proc. NAFIPS 90: Quater Century of Fuzziness, Vol.1, pp.9–12, 1990.Google Scholar