Fast convergence algorithm for wavelet neural network used for signal or function approximation

Author

Xiangyu, Song ; Feihu, Qi

Author_Institution

Dept. of Comput. Sci. & Eng., Shanghai Jiaotong Univ., China

Volume

2

fYear

1996

Firstpage

1401

Abstract

A new way to set the initial values of the wavelet neural network´s parameters is proposed in order to improve the convergence speed. Experiments on linear polynomials, exponent functions, sin & cos functions and a certain multistage simulation function show the neural network has a much faster convergence speed and can be widely used for approximating many kinds of signals and functions. A discussion on the merit of this method is given. The experiment results are satisfactory

Keywords

approximation theory; convergence of numerical methods; function approximation; neural nets; polynomials; signal processing; wavelet transforms; cos functions; experiment results; exponent functions; fast convergence algorithm; function approximation; initial values; linear polynomials; multistage simulation function; neural network parameters; signal approximation; sin functions; wavelet neural network; Artificial neural networks; Convergence; Discrete wavelet transforms; Frequency; Function approximation; Neural networks; Polynomials; Signal analysis; Wavelet analysis; Wavelet transforms;

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.1996.566584

Filename

566584