Deakin University
Browse

File(s) under embargo

Combining Monte Carlo tree search and apprenticeship learning for capture the flag

conference contribution
posted on 2024-01-09, 01:49 authored by J Ivanovo, WL Raffe, F Zambetta, X Li
In this paper we introduce a novel approach to agent control in competitive video games which combines Monte Carlo Tree Search (MCTS) and Apprenticeship Learning (AL). More specifically, an opponent model created through AL is used during the expansion phase of the Upper Confidence Bounds for Trees (UCT) variant of MCTS. We show how this approach can be applied to a game of Capture the Flag (CTF), an environment which is both non-deterministic and partially observable. The performance gain of a controller utilizing an opponent model learned via AL when compared to a controller using just UCT is shown both with win/loss ratios and True Skill rankings. Additionally, we build on previous findings by providing evidence of a bias towards a particular style of play in the AI Sandbox CTF environment. We believe that the approach highlighted here can be extended to a wider range of games other than just CTF.

History

Pagination

154-161

Location

TAIWAN, Tainan

Start date

2015-08-31

End date

2015-09-02

ISSN

2325-4270

ISBN-13

9781479986217

Language

English

Title of proceedings

2015 IEEE Conference on Computational Intelligence and Games, CIG 2015 - Proceedings

Event

IEEE Conference on Computational Intelligence and Games (CIG)

Publisher

IEEE

Place of publication

Piscataway, N.J.

Series

IEEE Conference on Computational Intelligence and Games

Usage metrics

    Research Publications

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC