A Review of the Application of Artificial Intelligence in Imperfect Information Games Represented by DouDiZhu
- DOI
- 10.2991/assehr.k.220701.033How to use a DOI?
- Keywords
- Imperfect Information Game; Monte Carlo Tree Search; Convolutional Neural Network; DouZero; DouDiZhu; Reinforcement Learning
- Abstract
Recently, deep reinforcement learning has achieved superhuman performance in various games such as Go, chess, and shogi. Compared with Go, DouDiZhu, also known as Chinese competitive poker, belongs to Imperfect Information Games (IIG), including hidden information, randomness, multi-agent cooperation, and competition. It is popular in China that it has become a national game. Imperfect information games, as an important branch of machine game, is closer to the decision-making under uncertainty in the complex real world than the perfect information games with transparent opponent information, and has a deeper research value. In addition, the game of DouDiZhu not only contains asymmetric information game, but also coexists competition and cooperation between players, so this relatively underdeveloped field is more considerable for research. This paper outlined the current mainstream reinforcement learning algorithms applied to this game and analyzed their respective advantages and disadvantages. This paper outlined the current mainstream reinforcement learning algorithms applied to this game and analyzed their respective advantages and disadvantages. At the same time, it prospected the future research directions.
- Copyright
- © 2022 The Authors. Published by Atlantis Press SARL.
- Open Access
- This is an open access article distributed under the CC BY-NC 4.0 license.
Cite this article
TY - CONF AU - Chuan He PY - 2022 DA - 2022/07/04 TI - A Review of the Application of Artificial Intelligence in Imperfect Information Games Represented by DouDiZhu BT - Proceedings of the 2022 International Conference on Science and Technology Ethics and Human Future (STEHF 2022) PB - Atlantis Press SP - 160 EP - 166 SN - 2352-5398 UR - https://doi.org/10.2991/assehr.k.220701.033 DO - 10.2991/assehr.k.220701.033 ID - He2022 ER -