An algorithm to determine neural network hidden layer size and weight coefficients

Author

Peng, Kemao ; Ge, Shuzhi S. ; Wen, Chuanyuan

Author_Institution

Dept. of Autom. Control, Beijing Univ. of Aeronaut. & Astronaut., China

fYear

2000

fDate

2000

Firstpage

261

Lastpage

266

Abstract

The strictly decreasing relationship between the sample approximation error and the number of hidden units in a three layer artificial feedforward neural network (AFNN) is proven in the sample space. The relationship is a powerful tool in determining the number of hidden units needed. A hybrid optimization algorithm is proposed on the relationship for simultaneously determining the number of hidden units and weight coefficients in the AFNN. The algorithm is the synthesis of golden section, evolutionary programming and gradient based algorithm which is effective in determining the number of hidden units and weight coefficients in the neural network

Keywords

feedforward neural nets; genetic algorithms; gradient methods; learning (artificial intelligence); evolutionary programming; feedforward neural network; gradient method; hidden units; learning; optimization; weight coefficients; Approximation error; Artificial neural networks; Ash; Convergence; Error analysis; Estimation theory; Genetic programming; Heuristic algorithms; Network synthesis; Neural networks;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligent Control, 2000. Proceedings of the 2000 IEEE International Symposium on

Conference_Location

Rio Patras

ISSN

2158-9860

Print_ISBN

0-7803-6491-0

Type

conf

DOI

10.1109/ISIC.2000.882934

Filename

882934