Determination of the appropriate node function of NNs by using the cascade-correlation algorithms

Weishui Wan; Kotaro Hirasawa; Junichi Murata; Chunzhi Jin; Jinglu Hu

Determination of the appropriate node function of NNs by using the cascade-correlation algorithms

Weishui Wan, Kotaro Hirasawa, Junichi Murata, Chunzhi Jin, Jinglu Hu

Section of Measurement and Control Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

How to determine the appropriate or optimal activation function in the neural networks for a specific learning samples remains open. In this paper the cascade-correlation algorithm which is an efficient constructive algorithm is used after implementation of some kinds of clustering algorithms to produce a modular network structure as a surrogate of activation node functions in the radial basis function (RBF) networks. In this way great improvement on the convergence rate of training algorithms and better approximation are achieved. Simulations with the two-spiral data sets proved the above assertion.

Original language	English
Title of host publication	IECON Proceedings (Industrial Electronics Conference)
Pages	1177-1182
Number of pages	6
Volume	2
Publication status	Published - 2000
Event	26th Annual Conference of the IEEE Electronics Society IECON 2000 - Nagoya, Japan Duration: Oct 22 2000 → Oct 28 2000

Other

Other	26th Annual Conference of the IEEE Electronics Society IECON 2000
Country/Territory	Japan
City	Nagoya
Period	10/22/00 → 10/28/00

All Science Journal Classification (ASJC) codes

Control and Systems Engineering
Electrical and Electronic Engineering

Cite this

@inproceedings{e0f37275fc84457ca16ff0ad75b26b95,

title = "Determination of the appropriate node function of NNs by using the cascade-correlation algorithms",

abstract = "How to determine the appropriate or optimal activation function in the neural networks for a specific learning samples remains open. In this paper the cascade-correlation algorithm which is an efficient constructive algorithm is used after implementation of some kinds of clustering algorithms to produce a modular network structure as a surrogate of activation node functions in the radial basis function (RBF) networks. In this way great improvement on the convergence rate of training algorithms and better approximation are achieved. Simulations with the two-spiral data sets proved the above assertion.",

author = "Weishui Wan and Kotaro Hirasawa and Junichi Murata and Chunzhi Jin and Jinglu Hu",

year = "2000",

language = "English",

volume = "2",

pages = "1177--1182",

booktitle = "IECON Proceedings (Industrial Electronics Conference)",

note = "26th Annual Conference of the IEEE Electronics Society IECON 2000 ; Conference date: 22-10-2000 Through 28-10-2000",

}

TY - GEN

T1 - Determination of the appropriate node function of NNs by using the cascade-correlation algorithms

AU - Wan, Weishui

AU - Hirasawa, Kotaro

AU - Murata, Junichi

AU - Jin, Chunzhi

AU - Hu, Jinglu

PY - 2000

Y1 - 2000

N2 - How to determine the appropriate or optimal activation function in the neural networks for a specific learning samples remains open. In this paper the cascade-correlation algorithm which is an efficient constructive algorithm is used after implementation of some kinds of clustering algorithms to produce a modular network structure as a surrogate of activation node functions in the radial basis function (RBF) networks. In this way great improvement on the convergence rate of training algorithms and better approximation are achieved. Simulations with the two-spiral data sets proved the above assertion.

AB - How to determine the appropriate or optimal activation function in the neural networks for a specific learning samples remains open. In this paper the cascade-correlation algorithm which is an efficient constructive algorithm is used after implementation of some kinds of clustering algorithms to produce a modular network structure as a surrogate of activation node functions in the radial basis function (RBF) networks. In this way great improvement on the convergence rate of training algorithms and better approximation are achieved. Simulations with the two-spiral data sets proved the above assertion.

UR - http://www.scopus.com/inward/record.url?scp=0034579408&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0034579408&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:0034579408

VL - 2

SP - 1177

EP - 1182

BT - IECON Proceedings (Industrial Electronics Conference)

T2 - 26th Annual Conference of the IEEE Electronics Society IECON 2000

Y2 - 22 October 2000 through 28 October 2000

ER -

Determination of the appropriate node function of NNs by using the cascade-correlation algorithms

Abstract

Other

All Science Journal Classification (ASJC) codes

Other files and links

Fingerprint

Cite this