Presentation | 2013-12-21 Difference of Enough Numbers for General and Regular Asymptotic Theories in Statistical Learning Sumio WATANABE, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | There are two asymptotic theories in statistical learning. One is the regular theory which assumes that the likelihood function can be approximated by a normal distribution, the other is the general theory which does not need such assumption. Prom the mathematical point of view, both theories need the limit condition that the number of traning samples goes to infinity. However, it strongly depends on the situation whether we can apply the theories to a practical problem or not. In this paper, we study the problem of discovery, and experimentally show that the regular theory can not be applied for observation of discovery of structure, whereas the general theory can be applicable. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | General asymptotic theory / Regular asymptotic theory / WAIC / Sample Number |
Paper # | NC2013-61 |
Date of Issue |
Conference Information | |
Committee | NC |
---|---|
Conference Date | 2013/12/14(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) | Difference of Enough Numbers for General and Regular Asymptotic Theories in Statistical Learning |
Sub Title (in English) | |
Keyword(1) | General asymptotic theory |
Keyword(2) | Regular asymptotic theory |
Keyword(3) | WAIC |
Keyword(4) | Sample Number |
1st Author's Name | Sumio WATANABE |
1st Author's Affiliation | Tokyo Institute of Technology() |
Date | 2013-12-21 |
Paper # | NC2013-61 |
Volume (vol) | vol.113 |
Number (no) | 374 |
Page | pp.pp.- |
#Pages | 6 |
Date of Issue |