Presentation | 1995/11/24 ARLNN: Layerd Neural Network with Autoregressive Links M. SEKIGUCHI, T. FURUYA, T. SATAKE, T. TANAKA, T. HIGUCHI, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | We propose a neural net (ARLNN) whose link has an autoregressive (AR) model. The AR model works as a delayed feedback loop which makes time sequence data processing possible. An leaning algorithm based on the backpropagation is shown. We applied the ARLNN to three spatio temporal pattern processing applications (growth estimation, abc trajectory, flipflop). ARLNN works well on those applications. The results show that the AR model between the input layer and the hidden layer plays important role in data processing. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Layerd Neural network / Auto-regressive Model / Backpropagation |
Paper # | NC95-65 |
Date of Issue |
Conference Information | |
Committee | NC |
---|---|
Conference Date | 1995/11/24(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) | ARLNN: Layerd Neural Network with Autoregressive Links |
Sub Title (in English) | |
Keyword(1) | Layerd Neural network |
Keyword(2) | Auto-regressive Model |
Keyword(3) | Backpropagation |
1st Author's Name | M. SEKIGUCHI |
1st Author's Affiliation | Toho University() |
2nd Author's Name | T. FURUYA |
2nd Author's Affiliation | Toho University |
3rd Author's Name | T. SATAKE |
3rd Author's Affiliation | University of Tsukuba |
4th Author's Name | T. TANAKA |
4th Author's Affiliation | Electrotechnical Laboratory |
5th Author's Name | T. HIGUCHI |
5th Author's Affiliation | Electrotechnical Laboratory |
Date | 1995/11/24 |
Paper # | NC95-65 |
Volume (vol) | vol.95 |
Number (no) | 389 |
Page | pp.pp.- |
#Pages | 8 |
Date of Issue |