Presentation | 2001/12/14 On-Line Learning of a Simple Perceptron Learning with Margin Kazuyuki HARA, Masato OKADA, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | We analyze a learning method using the margin κ a la Gardner for simple perceptron. The proposed method corresponds to the perceptron learning when κ=0, and it converges to the Hebbian learning when κ→∞. Neverthless, we found that the generalization ability of the present method outperforms those of the perceptron and the Hebbian methods at early stage of learning. Asymptotic property of the learning curve of the present method is analyzed by the computer simulation and it was the same as that of the perceptron learning. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | On-line learning / Margin / Simple perceptron / Generalization ability / Perceptron learning |
Paper # | NC2001-77 |
Date of Issue |
Conference Information | |
Committee | NC |
---|---|
Conference Date | 2001/12/14(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) | On-Line Learning of a Simple Perceptron Learning with Margin |
Sub Title (in English) | |
Keyword(1) | On-line learning |
Keyword(2) | Margin |
Keyword(3) | Simple perceptron |
Keyword(4) | Generalization ability |
Keyword(5) | Perceptron learning |
1st Author's Name | Kazuyuki HARA |
1st Author's Affiliation | Department of Electronics and Information Engineering, Tokyo Metropolitan College of Technology() |
2nd Author's Name | Masato OKADA |
2nd Author's Affiliation | Laboratry for Mathematical Neuroscience, Brain Science Institute, RIKEN |
Date | 2001/12/14 |
Paper # | NC2001-77 |
Volume (vol) | vol.101 |
Number (no) | 534 |
Page | pp.pp.- |
#Pages | 7 |
Date of Issue |