Simultaneous approximations of multivariate functions and their derivatives by neural networks with one hidden layer

作者: Xin Li

DOI: 10.1016/0925-2312(95)00070-4

关键词: Activation functionArtificial neural networkOpen setFunction (mathematics)Compact spaceDiscrete mathematicsOrder (group theory)MathematicsCombinatoricsApproximations of πPartial derivative

摘要: Abstract Let σ be a non-polynomial activation function of neural network that has nth order continuous derivatives on R. The objective this paper is to show for any compact set K Rs, s ≥ 1, and multivariate f defined an open containing K, with one hidden layer can so constructed all its existing kth partial derivatives, k = (k1, …, ks) ϵ Z+s satisfying ∑i 1ski ≤ n, simultaneously uniformly approximated by the network.

参考文章(14)
J Skrzypek, W Karplus, Neural networks in vision and pattern recognition World Scientific. ,(1992) , 10.1142/1695
Steven F. Zornetzer, Joel Davis, Thomas McKenna, Single neuron computation Academic Press Professional, Inc.. ,(1992)
Charles K. Chui, J. D. Ward, Larry L. Schumaker, Approximation Theory V ,(1986)
Ken-Ichi Funahashi, On the approximate realization of continuous mappings by neural networks Neural Networks. ,vol. 2, pp. 183- 192 ,(1989) , 10.1016/0893-6080(89)90003-8
Charles K. Chui, Xin Li, Approximation by ridge functions and neural networks with one hidden layer Journal of Approximation Theory. ,vol. 70, pp. 131- 141 ,(1992) , 10.1016/0021-9045(92)90081-X
Kurt Hornik, Maxwell Stinchcombe, Halbert White, Universal approximation of an unknown mapping and its derivatives using multilayer feedforward networks Neural Networks. ,vol. 3, pp. 551- 560 ,(1990) , 10.1016/0893-6080(90)90005-6
H.N Mhaskar, Charles A Micchelli, Approximation by superposition of sigmoidal and radial basis functions Advances in Applied Mathematics. ,vol. 13, pp. 350- 373 ,(1992) , 10.1016/0196-8858(92)90016-P
Michael J. D. Powell, Approximation theory and methods ,(1981)