Centre Selection for Radial Basis Function Networks
This paper is concerned with radial basis function neural networks. A new method is described by which means radial basis function centres can be selected. The method is based on a mean tracking clustering algorithm and it is shown how the approach provides a good procedure for radial basis function centre selection. Results are given which indicate how the method realises a network with excellent approximation properties.
KeywordsRadial Basis Function Radial Basis Function Neural Network Radial Basis Function Network Search Window Radial Basis Function Centre
Unable to display preview. Download preview PDF.
- Narendra, K.S.: Adaptive control of Dynamica Systems using Neural Networks, in Handbook of Intelligent Control, van Nostrand 1994.Google Scholar
- Sutanto, E., Warwick, K.: Proc. Cybernetics and Systems, 1, ed. Robert Trappl, 327 (1994).Google Scholar
- Craddock, R.J., Mason, J.D., Warwick, K: Proc IMACS Int. Sympòsium on Signal Processing Robotics and Neural Networks, 302 (1994).Google Scholar
- Soderstrom, T., Stoica, P.: System Identification. Prentice-Hall (1989).Google Scholar
- Vogt, W., Nagel, D.: Clinical Chemistry, 38, (1992).Google Scholar
- Sutanto, E.L.: The Design and Development of Multivariable Cluster Analysis for High-Speed Machinery. PhD Thesis, University of Reading (1995).Google Scholar