Nonlinear function approximation using radial basis function neural networks

Hafizah Husain, M. Khalid, R. Yusof

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

Radial basis function neural networks (RBFNN) which are best suited for nonlinear function approximation, have been successfully applied to a wide range of areas including system modeling. The two-stage training procedure adapted in numerous RBFNN applications usually provides satisfactory network performance. Though this method is proven to allow faster training and improves convergence, the initial stage of selecting the network centers pose a problem of creating a larger architecture than what is required. This limitation holds true in applications with large data samples. Various techniques have been developed to choose a sufficient number of centers to suit the network structure. Orthogonal least squares and input clustering are two of such methods that show considerable results of which can provide an amicable solution to the above problem. This paper presents a comparative study on the performance achieved by the two techniques demonstrated when applying the RBFNN in modeling of nonlinear functions and an investigation based on their capabilities in handling over-parameterization problems.

Original languageEnglish
Title of host publication2002 Student Conference on Research and Development: Globalizing Research and Development in Electrical and Electronics Engineering, SCOReD 2002 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages326-329
Number of pages4
ISBN (Electronic)0780375653, 9780780375659
DOIs
Publication statusPublished - 2002
Externally publishedYes
EventStudent Conference on Research and Development, SCOReD 2002 - Shah Alam, Malaysia
Duration: 16 Jul 200217 Jul 2002

Other

OtherStudent Conference on Research and Development, SCOReD 2002
CountryMalaysia
CityShah Alam
Period16/7/0217/7/02

Fingerprint

neural network
Neural networks
Network performance
Parameterization
performance

ASJC Scopus subject areas

  • Biomedical Engineering
  • Control and Systems Engineering
  • Electrical and Electronic Engineering
  • Education

Cite this

Husain, H., Khalid, M., & Yusof, R. (2002). Nonlinear function approximation using radial basis function neural networks. In 2002 Student Conference on Research and Development: Globalizing Research and Development in Electrical and Electronics Engineering, SCOReD 2002 - Proceedings (pp. 326-329). [1033124] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/SCORED.2002.1033124

Nonlinear function approximation using radial basis function neural networks. / Husain, Hafizah; Khalid, M.; Yusof, R.

2002 Student Conference on Research and Development: Globalizing Research and Development in Electrical and Electronics Engineering, SCOReD 2002 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2002. p. 326-329 1033124.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Husain, H, Khalid, M & Yusof, R 2002, Nonlinear function approximation using radial basis function neural networks. in 2002 Student Conference on Research and Development: Globalizing Research and Development in Electrical and Electronics Engineering, SCOReD 2002 - Proceedings., 1033124, Institute of Electrical and Electronics Engineers Inc., pp. 326-329, Student Conference on Research and Development, SCOReD 2002, Shah Alam, Malaysia, 16/7/02. https://doi.org/10.1109/SCORED.2002.1033124
Husain H, Khalid M, Yusof R. Nonlinear function approximation using radial basis function neural networks. In 2002 Student Conference on Research and Development: Globalizing Research and Development in Electrical and Electronics Engineering, SCOReD 2002 - Proceedings. Institute of Electrical and Electronics Engineers Inc. 2002. p. 326-329. 1033124 https://doi.org/10.1109/SCORED.2002.1033124
Husain, Hafizah ; Khalid, M. ; Yusof, R. / Nonlinear function approximation using radial basis function neural networks. 2002 Student Conference on Research and Development: Globalizing Research and Development in Electrical and Electronics Engineering, SCOReD 2002 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2002. pp. 326-329
@inproceedings{213f8f36f5df4c849515db3a16fda7a8,
title = "Nonlinear function approximation using radial basis function neural networks",
abstract = "Radial basis function neural networks (RBFNN) which are best suited for nonlinear function approximation, have been successfully applied to a wide range of areas including system modeling. The two-stage training procedure adapted in numerous RBFNN applications usually provides satisfactory network performance. Though this method is proven to allow faster training and improves convergence, the initial stage of selecting the network centers pose a problem of creating a larger architecture than what is required. This limitation holds true in applications with large data samples. Various techniques have been developed to choose a sufficient number of centers to suit the network structure. Orthogonal least squares and input clustering are two of such methods that show considerable results of which can provide an amicable solution to the above problem. This paper presents a comparative study on the performance achieved by the two techniques demonstrated when applying the RBFNN in modeling of nonlinear functions and an investigation based on their capabilities in handling over-parameterization problems.",
author = "Hafizah Husain and M. Khalid and R. Yusof",
year = "2002",
doi = "10.1109/SCORED.2002.1033124",
language = "English",
pages = "326--329",
booktitle = "2002 Student Conference on Research and Development: Globalizing Research and Development in Electrical and Electronics Engineering, SCOReD 2002 - Proceedings",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Nonlinear function approximation using radial basis function neural networks

AU - Husain, Hafizah

AU - Khalid, M.

AU - Yusof, R.

PY - 2002

Y1 - 2002

N2 - Radial basis function neural networks (RBFNN) which are best suited for nonlinear function approximation, have been successfully applied to a wide range of areas including system modeling. The two-stage training procedure adapted in numerous RBFNN applications usually provides satisfactory network performance. Though this method is proven to allow faster training and improves convergence, the initial stage of selecting the network centers pose a problem of creating a larger architecture than what is required. This limitation holds true in applications with large data samples. Various techniques have been developed to choose a sufficient number of centers to suit the network structure. Orthogonal least squares and input clustering are two of such methods that show considerable results of which can provide an amicable solution to the above problem. This paper presents a comparative study on the performance achieved by the two techniques demonstrated when applying the RBFNN in modeling of nonlinear functions and an investigation based on their capabilities in handling over-parameterization problems.

AB - Radial basis function neural networks (RBFNN) which are best suited for nonlinear function approximation, have been successfully applied to a wide range of areas including system modeling. The two-stage training procedure adapted in numerous RBFNN applications usually provides satisfactory network performance. Though this method is proven to allow faster training and improves convergence, the initial stage of selecting the network centers pose a problem of creating a larger architecture than what is required. This limitation holds true in applications with large data samples. Various techniques have been developed to choose a sufficient number of centers to suit the network structure. Orthogonal least squares and input clustering are two of such methods that show considerable results of which can provide an amicable solution to the above problem. This paper presents a comparative study on the performance achieved by the two techniques demonstrated when applying the RBFNN in modeling of nonlinear functions and an investigation based on their capabilities in handling over-parameterization problems.

UR - http://www.scopus.com/inward/record.url?scp=84898722556&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84898722556&partnerID=8YFLogxK

U2 - 10.1109/SCORED.2002.1033124

DO - 10.1109/SCORED.2002.1033124

M3 - Conference contribution

SP - 326

EP - 329

BT - 2002 Student Conference on Research and Development: Globalizing Research and Development in Electrical and Electronics Engineering, SCOReD 2002 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

ER -