Levenberg marquardt algorithm for the training of type-2 fuzzy neuro systems with a novel type-2 fuzzy membership function

A new training approach based on the Levenberg-Marquardt algorithm is proposed for type-2 fuzzy neural networks. While conventional gradient descent algorithms use only the first order derivative, the proposed algorithm used in this paper benefits from the first and the second order derivatives which makes the training procedure faster. Besides, this approach is more robust than the other techniques that use the second order derivatives, e.g. Gauss-Newton's method. The training algorithm proposed is tested on the training of a type-2 fuzzy neural network used for the prediction of a chaotic Mackey-Glass time series. The results show that the learning algorithm proposed not only results in faster training but also in a better forecasting accuracy.

[1]  Jerry M. Mendel,et al.  Stability analysis of type-2 fuzzy systems , 2008, 2008 IEEE International Conference on Fuzzy Systems (IEEE World Congress on Computational Intelligence).

[2]  Wei Chen,et al.  A fuzzy pattern recognition method of radar signal based on neural network , 2010, 2010 Sixth International Conference on Natural Computation.

[3]  Robert Babuska,et al.  Efficient training algorithm for Takagi-Sugeno type neuro-fuzzy network , 2001, 10th IEEE International Conference on Fuzzy Systems. (Cat. No.01CH37297).

[4]  Woei Tan,et al.  Uncertain Rule-Based Fuzzy Logic Systems: Introduction and New Directions (Mendel, J.M.; 2001) [book review] , 2007, IEEE Computational Intelligence Magazine.

[5]  Sung-Suk Kim,et al.  Development of Quantum-Based Adaptive Neuro-Fuzzy Networks , 2010, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).

[6]  Francis T.K. Au,et al.  Acceleration of Levenberg-Marquardt training of neural networks with variable decay rate , 2003, Proceedings of the International Joint Conference on Neural Networks, 2003..

[7]  Mohammad Sadegh Shahidzadeh,et al.  Damage detection using a new regularization method with variable parameter , 2010 .

[8]  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).

[9]  Jin Zhao,et al.  The application of fuzzy neural network in ship course control system , 2010, 2010 Seventh International Conference on Fuzzy Systems and Knowledge Discovery.

[10]  Mojtaba Ahmadieh Khanesar,et al.  A novel type-2 fuzzy membership function: application to the prediction of noisy data , 2010, 2010 IEEE International Conference on Computational Intelligence for Measurement Systems and Applications.

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

[12]  Okyay Kaynak,et al.  A type-2 neuro-fuzzy system based on clustering and gradient techniques applied to system identification and channel equalization , 2011, Appl. Soft Comput..

[13]  Yong Peng,et al.  A Fuzzy Optimization Neural Network Model Using Second Order Information , 2009, 2009 Sixth International Conference on Fuzzy Systems and Knowledge Discovery.

[14]  W. TanW.,et al.  Uncertain Rule-Based Fuzzy Logic Systems , 2007 .

[15]  Jerry M. Mendel,et al.  Interval type-2 fuzzy logic systems: theory and design , 2000, IEEE Trans. Fuzzy Syst..

[16]  Oscar Castillo,et al.  An improved method for edge detection based on interval type-2 fuzzy logic , 2010, Expert Syst. Appl..

[17]  Liu Kun,et al.  Fuzzy neural networks pattern recognition method and its application in ultrasonic detection for bonding defect of thin composite materials , 2009, 2009 IEEE International Conference on Automation and Logistics.

[18]  Fei Wang,et al.  Gait pattern estimation for intelligent bionic leg , 2008, 2008 Chinese Control and Decision Conference.