A Study on Go Algorithm of Monte-Carlo Tree Search Including Tabu Search
| Project/Area Number |
25330441
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Entertainment and game informatics 1
|
|---|
| Research Institution | Aichi Institute of Technology |
|---|
Principal Investigator |
Itoh Masaru 愛知工業大学, 情報科学部, 教授 (80221026)
|
|---|
| Project Period (FY) |
2013-04-01 – 2016-03-31
|
| Project Status |
Completed (Fiscal Year 2015)
|
| Budget Amount *help |
¥3,380,000 (Direct Cost: ¥2,600,000、Indirect Cost: ¥780,000)
Fiscal Year 2015: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2014: ¥1,690,000 (Direct Cost: ¥1,300,000、Indirect Cost: ¥390,000)
Fiscal Year 2013: ¥650,000 (Direct Cost: ¥500,000、Indirect Cost: ¥150,000)
|
|---|
| Keywords | コンピュータ囲碁 / モンテカルロ木探索 / プレイアウト / タブーリスト / 多様性 / ゾブリストハッシュ / 局面タブーリスト / 逐次更新法 / 一括更新法 / タブーサーチ的手法 / 詰碁 / 最良優先探索 / 幅優先探索 |
|---|
| Outline of Final Research Achievements |
The recently major algorithm in computer go is based on the Monte-Carlo tree search. The method needs about ten thousand playouts to obtain the next move in go playing. Here, playout is to be finishing the game from a given phase through the random moves. An improvement in the precision of playout could reduce the required number of playouts, and could form the more accurate Monte-Carlo search tree. However, that is surprisingly hard.
The research project has proposed that a playout has not accuracy but diversity in order to try to search more states for the current game phase. The diversity of playout restrains the growth of the tree toward the depth direction, and promotes the growth of it toward the uniformed breadth direction. The new method tries to introduce tabu lists, which is used as a short-term memory in the tabu search algorithm, into all playouts to carry out the diversity.
|
Report
(4 results)
Research Products
(13 results)
