A comparative study of learning methods in tuning parameters of fuzzy membership functions

We compare several popular training algorithms for tuning parameters of fuzzy membership functions (MFs). The algorithms compared are gradient descent (GD), resilient propagation (RPROP), Quickprop (QP), and Levenberg-Marquardt (LM) algorithms. These algorithms are combined with RLSE (recursive least squares estimate) to improve the efficiency of an ANFIS (adaptive network-based fuzzy inference system). The results, on average, show that the relative performance of these algorithms depends on the given task, but that RPROP produces better performance in terms of convergence speed, stability, and generalization properties.

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

[2]  Mohammad Bagher Menhaj,et al.  Training feedforward networks with the Marquardt algorithm , 1994, IEEE Trans. Neural Networks.

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

[4]  Mu-Song Chen,et al.  Fuzzy clustering analysis for optimizing fuzzy membership functions , 1999, Fuzzy Sets Syst..

[5]  Michio Sugeno,et al.  Fuzzy identification of systems and its applications to modeling and control , 1985, IEEE Transactions on Systems, Man, and Cybernetics.

[6]  L X Wang,et al.  Fuzzy basis functions, universal approximation, and orthogonal least-squares learning , 1992, IEEE Trans. Neural Networks.

[7]  Simon Haykin,et al.  Neural Networks: A Comprehensive Foundation , 1998 .

[8]  Scott E. Fahlman,et al.  An empirical study of learning speed in back-propagation networks , 1988 .

[9]  L. Glass,et al.  Oscillation and chaos in physiological control systems. , 1977, Science.

[10]  J. Lu,et al.  An on-line identification algorithm for fuzzy systems , 1994 .

[11]  Chin-Teng Lin,et al.  Neural-Network-Based Fuzzy Logic Control and Decision System , 1991, IEEE Trans. Computers.

[12]  Kumpati S. Narendra,et al.  Identification and control of dynamical systems using neural networks , 1990, IEEE Trans. Neural Networks.

[13]  Graham C. Goodwin,et al.  Adaptive filtering prediction and control , 1984 .