A Modified CMAC Algorithm Based on Credit Assignment

被引：0
作者：
Lei Zhang
Qixin Cao
Jay Lee
Yanzheng Zhao
机构：
[1] Shanghai Jiao Tong University,Research Institute of Robotics
[2] University of Wisconsin-Milwaukee,Research Center of Intelligent Maintenance Systems
来源：
Neural Processing Letters | 2004年 / 20卷
关键词：
CMAC; convergence property; credit assignment; learning interference;
D O I：
暂无
中图分类号：
学科分类号：
摘要：
A Credit-Assignment CMAC (CA-CMAC) algorithm is proposed to reduce learning interference in conventional CMAC. In the proposed CA-CMAC, the error of the training sample distributed to the addressed memory cell is proportional to the cell's credibility, which is the inverse of the cell's activated times. The learning process of CA-CMAC is analyzed and conventional CMAC is proved to be a special case of CA-CMAC. Furthermore, the convergence properties of CA-CMAC both in batch learning and in incremental learning are investigated; meanwhile, the convergence theorems in the two learning schemes are obtained, respectively. Finally, simulations are carried out to testify the theorems and compare the performance of CA-CMAC with that of CMAC. Simulation results prove that CA-CMAC converges faster than conventional CMAC.
引用
收藏
页码：1 / 10
页数：9
相关论文
共 25 条
[1]  
Albus J. S.(1975)A New approach to manipulator control; The Cerebellar Model Articulation Controller (CMAC) Journal of Dynamic System, Measurement and Control 97 220-233
[2]  
Cembrano G.(1999)Dynamic control of a robot arm using CMAC neural networks Control Engineering Practice 5 485-492
[3]  
Wells G.(1997)Learning convergence of CMAC techniques IEEE Transaction on Neural Networks 8 1281-1292
[4]  
Sarda J.(1991)CMAC-based adaptive critic learning control IEEE Transaction Neural Network 2 530-535
[5]  
Lin C. S.(2002)Enhance the performance of CMAC neural network via fuzzy theory and credit apportionment Proceedings of the 2002 International Joint Conference on Neural networks 1 715-720
[6]  
Chiang C. T.(1990)Real-time dynamic control of an industrial manipulators using a neural network-based learning controllers IEEE Transaction Robot. Automat. 6 1-9
[7]  
Lin C. S.(1992)A comparison of five algorithms for the training of CMAC learning control systems Automatica 28 1027-1035
[8]  
Kim H.(2002)A hybrid maximum error algorithm with neighborhood training for CMAC Proceedings of the 2002 International Joint Conference on Neural Networks 1 165-170
[9]  
Lu H. C.(2003)Credit Assigned CMAC and it's application to online learning robust controllers IEEE Transaction on Systems, Man, and Cybernetics-part B: Cybernetics 33 202-213
[10]  
Chang J.C.(1995)Neighborhood sequential and random training techniques for CAMC IEEE Trans Neural Network 6 196-202
123