Abstract
The approximation properties of the RBF neural networks are investigated in this paper. A new approach is proposed, which is based on approximations with orthogonal combinations of functions. An orthogonalization framework is presented for the Gaussian basis functions. It is shown how to use this framework to design efficient neural networks. Using this method we can estimate the necessary number of the hidden nodes, and we can evaluate how appropriate the use of the Gaussian RBF networks is for the approximation of a given function.
Similar content being viewed by others
References
A.R. Barron, “Universal approximation bounds for superpositions of a sigmoidal function”, IEEE Transactions on Information Theory, Vol. 39, pp. 930–945, 1993.
A.R. Barron, “Approximation andestimation bounds for artificial neural networks”, Machine Learning, Vol. 14, pp. 115–133, 1994.
G. Cybenko, “Approximation by superpositions of sigmoidal function”, Mathematics of Control, Signals, and Systems, Vol. 2, pp. 303–314, 1989.
S.W. Ellacott, “Aspects of numerical analysis of neural networks”, Acta Numerica, pp. 145–202, 1994.
S. Haykin,Neural Networks. A Comprehensive Foundation, Macmillan Publishing Company, New York, 1994.
K. Hornik, “Multilayer feedforward networks are universal approximators”, Neural Networks, Vol. 2, pp. 183–192, 1989.
K. Hornik, M. Stinchcombe, H. White and P. Auer, “Degree of approximation results for feedforward networks approximating unknown mappings and their derivatives”, Neural Computation, Vol. 6, pp. 1262–1275, 1994.
B. Irie and S. Miyake, “Capabilities of three-layered perceptrons”, Proceedings of IEEE-ICNN '88, pp. 641–648, 1988.
V. Kurková, P.C. Kainen and V. Kreinovich, “Estimates of the number of hidden units and variation with respect to half-spaces”, Neural Networks, Vol. 10, pp. 1061–1068, 1997.
N. Murata, “An integral representation offunctions using three-layered networks and their approximation bounds”, Neural Networks, Vol. 9, pp. 947–956, 1996.
T. Poggio and F. Girosi, “Networks forapproximation and learning”, Proc. IEEE, Vol. 78, pp. 1481–1497, 1990.
W. Rudin, Principles ofMathematical Analysis, Mûszaki Könyvkiadó, Budapest, 1978.
J. Park and I.W. Sandberg, “Universal approximationusing radial-basis function”, Neural Computation, Vol. 3, pp. 246–257, 1991.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
András, P. Orthogonal RBF Neural Network Approximation. Neural Processing Letters 9, 141–151 (1999). https://doi.org/10.1023/A:1018621308457
Issue Date:
DOI: https://doi.org/10.1023/A:1018621308457