Presentation | 2000/12/1 A method for merging hidden units of RBF networks without relearning of sample patterns Nobuhiko YAMAGUCHI, Koichiro YAMAUCHI, Naohiro ISHII, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | It is well known that we must setup appropriate number of hidden units for neural networks to make the generalization ability of the neural network maximum. In this paper, we propose a method for reducing the number of hidden units for RBF metworks, which has already finished the learning of sample patterns. In this method, the RBF network dose not relearn the sample patterns during the reduction process, so this method can be applied to several fields, where the system has no capacity to store the sample patterns. For example, this system can be used for reducing the number of hidden units of the network which has learned an environment by reinforcement learning. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Structual Learning / Merge / GRBF / Neural Networks |
Paper # | NC2000-75 |
Date of Issue |
Conference Information | |
Committee | NC |
---|---|
Conference Date | 2000/12/1(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) | A method for merging hidden units of RBF networks without relearning of sample patterns |
Sub Title (in English) | |
Keyword(1) | Structual Learning |
Keyword(2) | Merge |
Keyword(3) | GRBF |
Keyword(4) | Neural Networks |
1st Author's Name | Nobuhiko YAMAGUCHI |
1st Author's Affiliation | Department of Intelligence and Computer Science, Nagoya Institute of Technology() |
2nd Author's Name | Koichiro YAMAUCHI |
2nd Author's Affiliation | Information, Electronics and Systems Engineering, Graduate School of Engineering, Hokkaido University |
3rd Author's Name | Naohiro ISHII |
3rd Author's Affiliation | Department of Intelligence and Computer Science, Nagoya Institute of Technology |
Date | 2000/12/1 |
Paper # | NC2000-75 |
Volume (vol) | vol.100 |
Number (no) | 490 |
Page | pp.pp.- |
#Pages | 6 |
Date of Issue |