Explanation of positional evaluation by computer programs
| Project/Area Number |
25330432
|
|---|
| 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 | The University of Tokyo |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2013-04-01 – 2016-03-31
|
| Project Status |
Completed (Fiscal Year 2015)
|
| Budget Amount *help |
¥2,600,000 (Direct Cost: ¥2,000,000、Indirect Cost: ¥600,000)
Fiscal Year 2015: ¥650,000 (Direct Cost: ¥500,000、Indirect Cost: ¥150,000)
Fiscal Year 2014: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
Fiscal Year 2013: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
|
|---|
| Keywords | ゲームプログラミング / 探索 / 機械学習 / ゲーム情報学 / 人工知能 |
|---|
| Outline of Final Research Achievements |
The ultimate goal of this research is to present backgrounds and reasons of decisions made by computer programs, in a human-friendly manner. We focus on games such as shogi or go, because computer programs enhanced by game tree search and machine learning techniques are able to play decently as well as top-level human players do. We have developed a method for estimating relevance of possible future positions to be commented and improved its performance where the effectiveness of the method is evaluated through actual game records.
|
Report
(4 results)
Research Products
(21 results)
