Automatic fusion and splitting of artificial neural elements in optimizing the network size

Author

Kameyama, Keisuke ; Kosugi, Yukio

Author_Institution

Tokyo Inst. of Technol., Yokohama, Japan

fYear

1991

fDate

13-16 Oct 1991

Firstpage

1633

Abstract

A three-layered neural network that optimally self-adjusts the number of hidden layer units is proposed. The network combines two techniques: (1) unit fusion which enables an efficient pruning of the redundant units: and (2) linear transformations applied to the chosen hidden layer unit pair output and a modified backpropagation training rule for gradual fusion to reduce pruning shocks. The network was applied to a character recognition problem and it adjusted itself to a minimal configuration at high rate

Keywords

character recognition; learning systems; neural nets; optimisation; self-adjusting systems; backpropagation training rule; character recognition; fusion; hidden layer self adjusting; neural element splitting; optimisation; pruning; three-layered neural network; Artificial neural networks; Character recognition; Computer networks; Electric shock; Intelligent networks; Mutual information; Neural networks; Pattern classification; Problem-solving; Vectors;

fLanguage

English

Publisher

ieee

Conference_Titel

Systems, Man, and Cybernetics, 1991. 'Decision Aiding for Complex Systems, Conference Proceedings., 1991 IEEE International Conference on

Conference_Location

Charlottesville, VA

Print_ISBN

0-7803-0233-8

Type

conf

DOI

10.1109/ICSMC.1991.169926

Filename

169926