Presentation | 2020-03-05 [Memorial Lecture] A Tuning-Free Hardware Reservoir Based on MOSFET Crossbar Array for Practical Echo State Network Implementation Yuki Kume, Song Bian, Takashi Sato, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | Echo state network (ESN) is a class of recurrent neural network, and is known for drastically reducing the training time by the use of reservoir, a random and fixed network as the input and middle layers. In this paper, we propose a hardware implementation of ESN that uses practical MOSFET-based reservoir. As opposed to existing reservoirs that require additional tuning of network weights for improved stability, ourESN requires no post-training parameter tuning. To this end, we apply the circular law of random matrix to sparse reservoirs to determine a stable and fixed feedback gain. Through the evaluations using Mackey-Glass time-series dataset, the proposed ESN performs successful inference without post parameter tuning. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | echo state network / reservoir computing / recurrent neural network / MOSFET / crossbar array |
Paper # | VLD2019-118,HWS2019-91 |
Date of Issue | 2020-02-26 (VLD, HWS) |
Conference Information | |
Committee | HWS / VLD |
---|---|
Conference Date | 2020/3/4(4days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | Okinawa Ken Seinen Kaikan |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | Design Technology for System-on-Silicon, Hardware Security, etc. |
Chair | Shinichi Kawamura(Toshiba) / Nozomu Togawa(Waseda Univ.) |
Vice Chair | Makoto Ikeda(Univ. of Tokyo) / Yasuhisa Shimazaki(Renesas Electronics) / Daisuke Fukuda(Fujitsu Labs.) |
Secretary | Makoto Ikeda(SECOM) / Yasuhisa Shimazaki(Kyushu Univ.) / Daisuke Fukuda(Univ. of Aizu) |
Assistant | / Kazuki Ikeda(Hitachi) |
Paper Information | |
Registration To | Technical Committee on Hardware Security / Technical Committee on VLSI Design Technologies |
---|---|
Language | ENG |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | [Memorial Lecture] A Tuning-Free Hardware Reservoir Based on MOSFET Crossbar Array for Practical Echo State Network Implementation |
Sub Title (in English) | |
Keyword(1) | echo state network |
Keyword(2) | reservoir computing |
Keyword(3) | recurrent neural network |
Keyword(4) | MOSFET |
Keyword(5) | crossbar array |
1st Author's Name | Yuki Kume |
1st Author's Affiliation | Kyoto University(Kyoto Univ.) |
2nd Author's Name | Song Bian |
2nd Author's Affiliation | Kyoto University(Kyoto Univ.) |
3rd Author's Name | Takashi Sato |
3rd Author's Affiliation | Kyoto University(Kyoto Univ.) |
Date | 2020-03-05 |
Paper # | VLD2019-118,HWS2019-91 |
Volume (vol) | vol.119 |
Number (no) | VLD-443,HWS-444 |
Page | pp.pp.139-144(VLD), pp.139-144(HWS), |
#Pages | 6 |
Date of Issue | 2020-02-26 (VLD, HWS) |