| Presentation | 2011-01-24 Parameterized online quasi-Newton Training Algorithm for Feedforward Neural Networks Hiroshi NINOMIYA, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | This paper describes a new gradient based technique for training of feedforward neural networks. Recently, improved online quasi-Newton method was developed for neural network training improving feeding method of training data. This paper proposes a novel training algorithm based on the online quasi-Newton in which the feeding method of training data is parameterized. Furthermore, an analogy between the proposed algorithm and Langevin one is considered. The proposed algorithm is employed for robust neural network training purpose. Neural network training for some benchmark problems is presented to demonstrate the proposed algorithm. The proposed algorithm achieves more accurate and robust training results than the other quasi-Newton based training algorithms. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | feedforward neural networks / quasi-Newton method / online gradient training method / batch gradient training method / Langevin algorithm |
| Paper # | NLP2010-125,NC2010-89 |
| Date of Issue |
| Conference Information | |
| Committee | NC |
|---|---|
| Conference Date | 2011/1/17(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) | Parameterized online quasi-Newton Training Algorithm for Feedforward Neural Networks |
| Sub Title (in English) | |
| Keyword(1) | feedforward neural networks |
| Keyword(2) | quasi-Newton method |
| Keyword(3) | online gradient training method |
| Keyword(4) | batch gradient training method |
| Keyword(5) | Langevin algorithm |
| 1st Author's Name | Hiroshi NINOMIYA |
| 1st Author's Affiliation | Department of Information Science, Faculty of Engineering, Shonan Institute of Technology() |
| Date | 2011-01-24 |
| Paper # | NLP2010-125,NC2010-89 |
| Volume (vol) | vol.110 |
| Number (no) | 388 |
| Page | pp.pp.- |
| #Pages | 6 |
| Date of Issue | |