|
JAIST Repository >
School of Transdisciplinary Science >
Articles >
Journal Articles >
Please use this identifier to cite or link to this item:
https://hdl.handle.net/10119/20010
|
| Title: | More Human-Like Gameplay by Blending Policies From Supervised and Reinforcement Learning |
| Authors: | Ogawa, Tatsuyoshi
Hsueh, Chu-Hsuan
Ikeda, Kokolo |
| Keywords: | Board game
human-likeness
player modeling
reinforcement learning
supervised learning |
| Issue Date: | 2024-07-11 |
| Publisher: | Institute of Electrical and Electronics Engineers (IEEE) |
| Magazine name: | IEEE Transactions on Games |
| Volume: | 16 |
| Number: | 4 |
| Start page: | 831 |
| End page: | 843 |
| DOI: | 10.1109/TG.2024.3424668 |
| Abstract: | Modeling human players’ behaviors in games is a key challenge for making natural computer players, evaluating games, and generating content. To achieve better human–computer interaction, researchers have tried various methods to create human-like artificial intelligence. In chess and Go, supervised learning with deep neural networks is known as one of the most effective ways to predict humanmoves. However, formany other games (e.g., Shogi), it is hard to collect a similar amount of game records, resulting in poor move-matching accuracy of the supervised learning. We propose a method to compensate for the weakness of the supervised learning policy by Blending it with an AlphaZero-like reinforcement learning policy. Experiments on Shogi showed that the Blend method significantly improved the move-matching accuracy over supervised learning models. Experiments on chess and Go with a limited number of game records also showed similar results. The Blendmethodwas effectivewith bothmedium and large numbers of games, particularly the medium case.We confirmed the robustness of the Blend model to the parameter and discussed the mechanism why themove-matching accuracy improves. In addition,we showed that theBlend model performed better than existingwork that tried to improve the move-matching accuracy. |
| Rights: | Copyright (c) 2024 Author(s). Tatsuyoshi Ogawa, Chu-Hsuan Hsueh, and Kokolo Ikeda, IEEE Transactions on Games, vol. 16, no. 4, pp. 831-843. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). Original publication is available on IEEE Xplore via https://doi.org/10.1109/TG.2024.3424668. |
| URI: | https://hdl.handle.net/10119/20010 |
| Material Type: | publisher |
| Appears in Collections: | d10-1. 雑誌掲載論文 (Journal Articles)
|
Files in This Item:
| File |
Description |
Size | Format |
|---|
| T-IKEDA-K-0930-2.pdf | | 804Kb | Adobe PDF | View/Open |
|
All items in DSpace are protected by copyright, with all rights reserved.
|
