A hybrid search agent in pommerman

Hongwei Zhou, Yichen Gong, Luvneesh Mugrai, Ahmed Khalifa, Andy Nealen, Julian Togelius

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Abstract

    In this paper, we explore the possibility of search-based agents in games with resource-intensive forward models. We implemented a player agent in the Pommerman framework and put it against the baseline agent to measure its performance. We implemented a heuristic agent and improved it by enabling depth-limited tree search in specific gameplay moments. We also compared different node selection methods during depth-limited tree search. Our result shows that depth-limited tree search is still viable when presented with inefficient forward models and exploitation-driven selection method is the most efficient in this specific domain.

    Original languageEnglish (US)
    Title of host publicationProceedings of the 13th International Conference on the Foundations of Digital Games, FDG 2018
    EditorsSebastian Deterding, Mitu Khandaker, Sebastian Risi, Jose Font, Steve Dahlskog, Christoph Salge, Carl Magnus Olsson
    PublisherAssociation for Computing Machinery
    ISBN (Electronic)9781450365710
    DOIs
    StatePublished - Aug 7 2018
    Event13th International Conference on the Foundations of Digital Games, FDG 2018 - Malmo, Sweden
    Duration: Aug 7 2018Aug 10 2018

    Publication series

    NameACM International Conference Proceeding Series

    Other

    Other13th International Conference on the Foundations of Digital Games, FDG 2018
    Country/TerritorySweden
    CityMalmo
    Period8/7/188/10/18

    Keywords

    • Monte carlo methods
    • Pommerman
    • State machines
    • Tree search

    ASJC Scopus subject areas

    • Software
    • Human-Computer Interaction
    • Computer Vision and Pattern Recognition
    • Computer Networks and Communications

    Cite this