论文信息 - Uniform Approximation by Neural Networks

Uniform Approximation by Neural Networks

LetD?Rdbe a compact set and let?be a uniformly bounded set ofD?Rfunctions. For a given real-valued functionfdefined onDand a given natural numbern, we are looking for a good uniform approximation tofof the form ?ni=1a1?i, with?i??,ai?R. Two main cases are considered: (1) whenDis a finite set and (2) when the set?is formed by the functions?v,b(x)?s(v·x+b), wherev?Rd,b?R, andsis a fixedR?Rfunction.

Y. Makovoz

[1] Michel Loève,et al. Probability Theory I , 1977 .

[2] H. Warren. Lower bounds for approximation by nonlinear manifolds , 1968 .

[3] P. Heywood. Trigonometric Series , 1968, Nature.

[4] Vladimir Vapnik,et al. Chervonenkis: On the uniform convergence of relative frequencies of events to their probabilities , 1971 .

[5] Y Makovoz. On trigonometric n-widths and their generalization , 1984 .

[6] Andrew R. Barron,et al. Universal approximation bounds for superpositions of a sigmoidal function , 1993, IEEE Trans. Inf. Theory.

[7] R. DeVore,et al. Nonlinear Approximation by Trigonometric Sums , 1995 .

[8] Halbert White,et al. Sup-norm approximation bounds for networks through probabilistic methods , 1995, IEEE Trans. Inf. Theory.

[9] Y. Makovoz. Random Approximants and Neural Networks , 1996 .

[10] G. Lorentz,et al. Constructive approximation : advanced problems , 1996 .

[11] E. S. Belinskii. Decomposition theorems and approximation by a “floating" system of exponentials , 1998 .