Separated Antecedent and Consequent Learning for Takagi-Sugeno Fuzzy Systems

In this paper a new algorithm for the learning of Takagi-Sugeno fuzzy systems is introduced. In the algorithm different learning techniques are applied for the antecedent and the consequent parameters of the fuzzy system. We propose a hybrid method for the antecedent parameters learning based on the combination of the bacterial evolutionary algorithm (BEA) and the Levenberg-Marquardt (LM) method. For the linear parameters in fuzzy systems appearing in the rule consequents the least squares (LS) and the recursive least squares (RLS) techniques are applied, which will lead to a global optimal solution of linear parameter vectors in the least squares sense. Therefore a better performance can be guaranteed than with a complete learning by BEA and LM. The paper is concluded by evaluation results based on high-dimensional test data. These evaluation results compare the new method with some conventional fuzzy training methods with respect to approximation accuracy and model complexity.

[1]  Lennart Ljung,et al.  System Identification: Theory for the User , 1987 .

[2]  P. Werbos,et al.  Beyond Regression : "New Tools for Prediction and Analysis in the Behavioral Sciences , 1974 .

[3]  John Yen,et al.  Improving the interpretability of TSK fuzzy models by combining global learning and local learning , 1998, IEEE Trans. Fuzzy Syst..

[4]  Jyh-Shing Roger Jang,et al.  ANFIS: adaptive-network-based fuzzy inference system , 1993, IEEE Trans. Syst. Man Cybern..

[5]  Dimitar Filev,et al.  Generation of Fuzzy Rules by Mountain Clustering , 1994, J. Intell. Fuzzy Syst..

[6]  Robert Babuska,et al.  Fuzzy Modeling for Control , 1998 .

[7]  Ulrich Bodenhofer,et al.  Incremental Learning of Fuzzy Basis Function Networks with a Modified Version of Vector Quantization , 2006 .

[8]  Chuen-Tsai Sun,et al.  Functional equivalence between radial basis function networks and fuzzy inference systems , 1993, IEEE Trans. Neural Networks.

[9]  D. Ruppert.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction , 2004 .

[10]  M. Stone.  Cross‐Validatory Choice and Assessment of Statistical Predictions , 1976 .

[11]  Takeshi Furuhashi,et al.  Fuzzy system parameters discovery by bacterial evolutionary algorithm , 1999, IEEE Trans. Fuzzy Syst..

[12]  Bart Kosko,et al.  Fuzzy Systems as Universal Approximators , 1994, IEEE Trans. Computers.

[13]  Michael Spann,et al.  A new approach to clustering , 1990, Pattern Recognit..

[14]  László T. Kóczy,et al.  Estimating fuzzy membership functions parameters by the Levenberg-Marquardt algorithm , 2004, 2004 IEEE International Conference on Fuzzy Systems (IEEE Cat. No.04CH37542).

[15]  László T. Kóczy,et al.  1 On functional equivalence of certain fuzzy controllers and RBF type approximation schemes ? , 2000 .

[16]  John H. Holland,et al.  Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence , 1992 .

[17]  Eyke Hüllermeier,et al.  Improving the interpretability of data-driven evolving fuzzy systems , 2005, EUSFLAT Conf..

[18]  D. Marquardt.  An Algorithm for Least-Squares Estimation of Nonlinear Parameters , 1963 .

[19]  Ulrich Bodenhofer,et al.  Tuning of Fuzzy Systems as an Ill-Posed Problem , 2002 .

[20]  Edwin Lughofer,et al.  FLEXFIS: A Variant for Incremental Learning of Takagi-Sugeno Fuzzy Systems , 2005, The 14th IEEE International Conference on Fuzzy Systems, 2005. FUZZ '05..

[21]  L.T. Koczy,et al.  Supervised training algorithms for B-spline neural networks and fuzzy systems , 2001, Proceedings Joint 9th IFSA World Congress and 20th NAFIPS International Conference (Cat. No. 01TH8569).