A new CMAC neural network model with adaptive quantization input layer

Author

Xiaozhi, Gao ; Changhong, Wang ; Gao, X.M. ; Ovaska, Seppo J.

Author_Institution

Dept. of Control Eng., Harbin Inst. of Technol., China

Volume

2

fYear

1996

Firstpage

1417

Abstract

We first discuss the structure, principle and learning algorithm of the cerebellar model arithmetic controller (CMAC) neural network model. A new adaptive quantization method based on competitive learning is then proposed to quantize the inputs of the CMAC according to the degree of variations of the approximated function. Theoretical analysis and simulation results show that with the input layer using this algorithm the CMAC can provide a more accurate and efficient approximation than the original model using equal-size quantization method

Keywords

adaptive signal processing; cerebellar model arithmetic computers; function approximation; quantisation (signal); unsupervised learning; CMAC neural network model; adaptive quantization input layer; approximated function; approximation; cerebellar model arithmetic controller; competitive learning; equal size quantization method; input layer; learning algorithm; simulation results; Adaptive systems; Algorithm design and analysis; Arithmetic; Control engineering; Function approximation; Laboratories; Neural networks; Quantization; Robots; Signal processing algorithms;

fLanguage

English

Publisher

ieee

Conference_Titel

Signal Processing, 1996., 3rd International Conference on

Conference_Location

Beijing

Print_ISBN

0-7803-2912-0

Type

conf

DOI

10.1109/ICSIGP.1996.566589

Filename

566589