Volume 5 Number 5 (Sep. 2016)
Home > Archive > 2016 > Volume 5 Number 5 (Sep. 2016) >
IJCCE 2016 Vol.5(5): 311-320 ISSN: 2010-3743
DOI: 10.17706/IJCCE.2016.5.5.311-320

Dynamic Threshold Selection for Sequential Learning in Radial Basis Function Networks

W. S. Lim, W. L. Yeoh
Abstract—For sequential learning in Radial Basis Function (RBF) Networks, there is a requirement for dynamic selection of threshold because a constant threshold is inadequate to accommodate functions of varying amplitudes. In this paper, a new criterion is defined for the dynamic selection of the Euclidean output deviation threshold. Its effect on the learning process experienced by RBF networks with regard to functions of variable amplitude is shown. This improved network can automatically select a suitable threshold for its own supervised learning depending on the objective parameters set to achieve certain accuracy level of the desired output. This paper also proposes further automation to neuron growing and pruning within Radial Basis Function (RBF) neural networks. The proposed dynamic threshold selection technique has shown significant improvement in achieving stable neuron growth rate in dealing with signal amplitude variation.

Index Terms—Dynamic threshold selection, radial basis function, neural network, sequential learning, nonlinear function approximation, minimal resource allocation network.

The authors are with Multimedia University, Faculty of Engineering & Technology, 75450 Melaka, Malaysia.

Cite:W. S. Lim, W. L. Yeoh, "Dynamic Threshold Selection for Sequential Learning in Radial Basis Function Networks," International Journal of Computer and Communication Engineering vol. 5, no. 5, pp. 311-320, 2016.

General Information

ISSN: 2010-3743
Frequency: Quarterly
Editor-in-Chief: Dr. Maode Ma
Abstracting/ Indexing: EI (INSPEC, IET), Google Scholar, Crossref, Engineering & Technology Digital Library, ProQuest, and Electronic Journals Library
E-mail: ijcce@iap.org
  • Mar 31, 2016 News!

    IJCCE Vol. 4, No. 5 has been indexed by EI (Inspec) Inspec, created by the Institution of Engineering and Tech.!   [Click]

  • Jun 16, 2017 News!

    Vol. 4, No. 5 has been indexed by EI (Inspec)!   [Click]

  • Jun 15, 2017 News!

    Vol. 4, No. 4 has been indexed by EI (Inspec)!   [Click]

  • Jun 14, 2017 News!

    ICCNT 2017 will be held in Zurich, Switzerland during September 1-3, 2017 ! Welcome to submit your paper to ICCNT 2017 !   [Click]

  • Jun 14, 2017 News!

    Vol. 4, No. 3 has been indexed by EI (Inspec)!   [Click]

  • Read more>>