Presentation | 1997/12/12 FAST BACK-PROPAGATION WITH AUTOMATICALLY ADJUSTABLE LEARNING RATE Peter GECZY, Shiro USUI, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | An effective novel approach to first order optimization techniques for training MLP artificial neural networks is presented. The approach is based on modified line search subproblem for back-propagation learning mechanism, where the learning rate is automatically adjusted at each iteration. The appropriate learning rate is obtained in a single calculation. This remarkably simplifies computational complexity of the line search subproblem. The proposed algorithm is built on solid theoretical grounds rather than heuristics. The algorithm is convergent, with superlinear convergence rates and linear computational complexity. Effectiveness of the algorithm is tested on five data sets. Results indicate superior performance of the newly proposed algorithm over the standard back-propagation technique. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | first order optimization / steepest descent / line search subproblem / adjustable learning rate |
Paper # | NC97-61 |
Date of Issue |
Conference Information | |
Committee | NC |
---|---|
Conference Date | 1997/12/12(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 | ENG |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | FAST BACK-PROPAGATION WITH AUTOMATICALLY ADJUSTABLE LEARNING RATE |
Sub Title (in English) | |
Keyword(1) | first order optimization |
Keyword(2) | steepest descent |
Keyword(3) | line search subproblem |
Keyword(4) | adjustable learning rate |
1st Author's Name | Peter GECZY |
1st Author's Affiliation | Department of Information and Computer Sciences, Toyohashi University of Technology() |
2nd Author's Name | Shiro USUI |
2nd Author's Affiliation | Department of Information and Computer Sciences, Toyohashi University of Technology |
Date | 1997/12/12 |
Paper # | NC97-61 |
Volume (vol) | vol.97 |
Number (no) | 448 |
Page | pp.pp.- |
#Pages | 8 |
Date of Issue |