Title :
Two-level high precision neural network
Author_Institution :
Inst. of Control Sci., Moscow
Abstract :
A brief description of the CMAC neural network and the CMAC mapping algorithm is given. The influence of noise on the accuracy of plant identification is shown. It is proposed to use a hierarchical structure of the CMAC with two levels in order to provide it with the filtering property which allows one to eliminate the noise influence completely for a stationary case. Numerical results are presented
Keywords :
cerebellar model arithmetic computers; filtering theory; function approximation; identification; learning (artificial intelligence); noise; CMAC mapping algorithm; CMAC neural network; filtering; function approximation; hierarchical structure; identification; learning process; Adaptive control; Associative memory; Books; Filtering; Manipulator dynamics; Neural networks; Noise level; Pattern recognition; Robot control; Table lookup;
Conference_Titel :
Control of Oscillations and Chaos, 1997. Proceedings., 1997 1st International Conference
Conference_Location :
St. Petersburg
Print_ISBN :
0-7803-4247-X
DOI :
10.1109/COC.1997.626659
