Presentation | 2015-09-18 Trax solver based on machine-learned evaluation function Takuya Nakamichi, Yusuke Sonoda, Takayuki Matsuzaki, Motoki Amagasaki, Masahiro Iida, Morihiro Kuga, Toshinori Sueyoshi, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | We develop a solver of board game Trax. Our basic strategy is a common game tree search algorithm. We explore the best move by the alpha-beta pruning. The game tree search is required evaluation function of some sort. we create it by machine learning. Usually this function is made by machine learning based on game record data of expert, but in this game, there is not enough game record data. Therefore, we make the evaluation function by machine learning based on the winning percentage by the Monte Carlo tree search. The search possible number of nodes is small in software-only solver. So we implement a partial of it on FPGA(Field Programmable Gate Array) for acceleration. The solver was about 10 times faster than a software-only solver. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Accelerator / Machine learning / Monte Carlo tree search / Trax / FPGA |
Paper # | RECONF2015-33 |
Date of Issue | 2015-09-11 (RECONF) |
Conference Information | |
Committee | RECONF |
---|---|
Conference Date | 2015/9/18(2days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | Ehime University |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | Reconfigurable Systems, etc. |
Chair | Minoru Watanabe(Shizuoka Univ.) |
Vice Chair | Masato Motomura(Hokkaido Univ.) / Yuichiro Shibata(Nagasaki Univ.) |
Secretary | Masato Motomura(Toshiba) / Yuichiro Shibata(Univ. of Tsukuba) |
Assistant | Kazuya Tanikagawa(Hiroshima City Univ.) / Takefumi Miyoshi(e-trees.Japan) |
Paper Information | |
Registration To | Technical Committee on Reconfigurable Systems |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Trax solver based on machine-learned evaluation function |
Sub Title (in English) | |
Keyword(1) | Accelerator |
Keyword(2) | Machine learning |
Keyword(3) | Monte Carlo tree search |
Keyword(4) | Trax |
Keyword(5) | FPGA |
1st Author's Name | Takuya Nakamichi |
1st Author's Affiliation | Kumamoto University(Kumamoto Univ.) |
2nd Author's Name | Yusuke Sonoda |
2nd Author's Affiliation | Kumamoto University(Kumamoto Univ.) |
3rd Author's Name | Takayuki Matsuzaki |
3rd Author's Affiliation | Kumamoto University(Kumamoto Univ.) |
4th Author's Name | Motoki Amagasaki |
4th Author's Affiliation | Kumamoto University(Kumamoto Univ.) |
5th Author's Name | Masahiro Iida |
5th Author's Affiliation | Kumamoto University(Kumamoto Univ.) |
6th Author's Name | Morihiro Kuga |
6th Author's Affiliation | Kumamoto University(Kumamoto Univ.) |
7th Author's Name | Toshinori Sueyoshi |
7th Author's Affiliation | Kumamoto University(Kumamoto Univ.) |
Date | 2015-09-18 |
Paper # | RECONF2015-33 |
Volume (vol) | vol.115 |
Number (no) | RECONF-228 |
Page | pp.pp.7-12(RECONF), |
#Pages | 6 |
Date of Issue | 2015-09-11 (RECONF) |