Presentation | 2011-07-25 Statistical Learning Theory for Unrealizable and Singular Cases Sumio WATANABE, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | Artificial neural networks are nonlinear and nonregular learning machines, whose learning performances can not been clarified by the conventional statistical theories. In this paper, we study the case that the optimal model is essentially nonunique for the true distribution, and show that the asymptotic behaviors of the free energy and the generalization error are different from the conventional ones. We also show that the spontaneous symmetry breaking of the posterior distribution occurs by the fluctuation of the training samples. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Neural Network / Learning Theorey / Spontaneous Symmetry Breaking |
Paper # | NC2011-24 |
Date of Issue |
Conference Information | |
Committee | NC |
---|---|
Conference Date | 2011/7/18(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 | Neurocomputing (NC) |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Statistical Learning Theory for Unrealizable and Singular Cases |
Sub Title (in English) | |
Keyword(1) | Neural Network |
Keyword(2) | Learning Theorey |
Keyword(3) | Spontaneous Symmetry Breaking |
1st Author's Name | Sumio WATANABE |
1st Author's Affiliation | Tokyo Institute of Technology, Dept. of Computational Intelligence and Systems Science() |
Date | 2011-07-25 |
Paper # | NC2011-24 |
Volume (vol) | vol.111 |
Number (no) | 157 |
Page | pp.pp.- |
#Pages | 6 |
Date of Issue |