Learning to Play Imperfect-Information Games by Imitating an Oracle Planner

Loading...
Thumbnail Image
Journal Title
Journal ISSN
Volume Title
A1 Alkuperäisartikkeli tieteellisessä aikakauslehdessä
Date
2022
Major/Subject
Mcode
Degree programme
Language
en
Pages
11
Series
IEEE Transactions on Games
Abstract
We consider learning to play multiplayer imperfect-information games with simultaneous moves and large state-action spaces. Previous attempts to tackle such challenging games have largely focused on model-free learning methods, often requiring hundreds of years of experience to produce competitive agents. Our approach is based on model-based planning. We tackle the problem of partial observability by first building an (oracle) planner that has access to the full state of the environment and then distilling the knowledge of the oracle to a (follower) agent which is trained to play the imperfect-information game by imitating the oracle's choices. We experimentally show that planning with naive Monte Carlo tree search performs poorly in large combinatorial action spaces. We therefore propose planning with a fixed-depth tree search and decoupled Thompson sampling for action selection. We show that the planner is able to discover efficient playing strategies in the games of Clash Royale and Pommerman and the follower policy successfully learns to implement them by training on few hundred battles.
Description
Publisher Copyright: IEEE Copyright: Copyright 2021 Elsevier B.V., All rights reserved.
Keywords
Buildings, Clash Royale, Games, Imperfect-information games, Law, Monte Carlo methods, Monte Carlo tree search, Observability, Planning, Poles and towers, Pommerman
Other note
Citation
Boney, R, Ilin, A, Kannala, J & Seppanen, J 2022, ' Learning to Play Imperfect-Information Games by Imitating an Oracle Planner ', IEEE Transactions on Games, vol. 14, no. 2, pp. 262-272 . https://doi.org/10.1109/TG.2021.3067723