Presentation | 1994/6/17 On Constructive Function Approximation by Three-Layer Artificial Neural Networks and its Approximation Error Specification Shin Suzuki, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | Function approximation is a key issue in evaluating the computational ability of multilayer artificial neural networks. Most previous studies showed only the existence of approximating networks by non-constructive methods,contributing almost nothing to the construction of approximating networks and the specification of the number of hidden-layer units and the approximation error.In this study,a three-layer artificial neural network with sigmoidal hidden-layer units approximating a 2π-perio dic pthorder Lebesgue integrable function on Rm to Rn is constructed,specifying the number of hidden-layer units and approximation error.This result can easily be applied to the approximation to pth-order Lebesgue integrable functions defined on a product set of bounded intervals in Rm to Rn.This study gives a new step to evaluate the computational ability of multilayer artificial neural networks. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | neural network / function approximation / Jackson′s theory / nu mber of hidden-layer units / error specification |
Paper # | NLP94-27 |
Date of Issue |
Conference Information | |
Committee | NLP |
---|---|
Conference Date | 1994/6/17(1days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | |
Vice Chair | |
Secretary | |
Assistant |
Paper Information | |
Registration To | Nonlinear Problems (NLP) |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | On Constructive Function Approximation by Three-Layer Artificial Neural Networks and its Approximation Error Specification |
Sub Title (in English) | |
Keyword(1) | neural network |
Keyword(2) | function approximation |
Keyword(3) | Jackson′s theory |
Keyword(4) | nu mber of hidden-layer units |
Keyword(5) | error specification |
1st Author's Name | Shin Suzuki |
1st Author's Affiliation | Basic Research Laboratories,NTT() |
Date | 1994/6/17 |
Paper # | NLP94-27 |
Volume (vol) | vol.94 |
Number (no) | 98 |
Page | pp.pp.- |
#Pages | 8 |
Date of Issue |