Hyper-heuristic general video game playing

Andre Mendes, Julian Togelius, Andrew Nealen

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

    Abstract

    In general video game playing, the challenge is to create agents that play unseen games proficiently. Stochastic tree search algorithms, like Monte Carlo Tree Search, perform relatively well on this task. However, performance is non-transitive: different agents perform best in different games, which means that there is not a single agent that is the best in all the games. Rather, some types of games are dominated by a few agents whereas other different agents dominate other types of games. Thus, it should be possible to construct a hyper-agent that selects from a portfolio, in which constituent sub-agents will play a new game best. Since there is no knowledge about the games, the agent needs to use available features to predict the most suitable algorithm. This work constructs such a hyper-agent using the General Video Game Playing Framework (GVGAI). The proposed method achieves promising results that show the applicability of hyper-heuristics in general video game playing and related tasks.

    Original languageEnglish (US)
    Title of host publication2016 IEEE Conference on Computational Intelligence and Games, CIG 2016
    PublisherIEEE Computer Society
    ISBN (Electronic)9781509018833
    DOIs
    StatePublished - Feb 21 2017
    Event2016 IEEE Conference on Computational Intelligence and Games, CIG 2016 - Santorini, Greece
    Duration: Sep 20 2016Sep 23 2016

    Other

    Other2016 IEEE Conference on Computational Intelligence and Games, CIG 2016
    CountryGreece
    CitySantorini
    Period9/20/169/23/16

    ASJC Scopus subject areas

    • Artificial Intelligence
    • Computer Graphics and Computer-Aided Design
    • Computer Vision and Pattern Recognition
    • Human-Computer Interaction
    • Software

    Cite this

    Mendes, A., Togelius, J., & Nealen, A. (2017). Hyper-heuristic general video game playing. In 2016 IEEE Conference on Computational Intelligence and Games, CIG 2016 [7860398] IEEE Computer Society. https://doi.org/10.1109/CIG.2016.7860398

    Hyper-heuristic general video game playing. / Mendes, Andre; Togelius, Julian; Nealen, Andrew.

    2016 IEEE Conference on Computational Intelligence and Games, CIG 2016. IEEE Computer Society, 2017. 7860398.

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

    Mendes, A, Togelius, J & Nealen, A 2017, Hyper-heuristic general video game playing. in 2016 IEEE Conference on Computational Intelligence and Games, CIG 2016., 7860398, IEEE Computer Society, 2016 IEEE Conference on Computational Intelligence and Games, CIG 2016, Santorini, Greece, 9/20/16. https://doi.org/10.1109/CIG.2016.7860398
    Mendes A, Togelius J, Nealen A. Hyper-heuristic general video game playing. In 2016 IEEE Conference on Computational Intelligence and Games, CIG 2016. IEEE Computer Society. 2017. 7860398 https://doi.org/10.1109/CIG.2016.7860398
    Mendes, Andre ; Togelius, Julian ; Nealen, Andrew. / Hyper-heuristic general video game playing. 2016 IEEE Conference on Computational Intelligence and Games, CIG 2016. IEEE Computer Society, 2017.
    @inproceedings{5ba41bacba7f4271af4b1d698b1ebff7,
    title = "Hyper-heuristic general video game playing",
    abstract = "In general video game playing, the challenge is to create agents that play unseen games proficiently. Stochastic tree search algorithms, like Monte Carlo Tree Search, perform relatively well on this task. However, performance is non-transitive: different agents perform best in different games, which means that there is not a single agent that is the best in all the games. Rather, some types of games are dominated by a few agents whereas other different agents dominate other types of games. Thus, it should be possible to construct a hyper-agent that selects from a portfolio, in which constituent sub-agents will play a new game best. Since there is no knowledge about the games, the agent needs to use available features to predict the most suitable algorithm. This work constructs such a hyper-agent using the General Video Game Playing Framework (GVGAI). The proposed method achieves promising results that show the applicability of hyper-heuristics in general video game playing and related tasks.",
    author = "Andre Mendes and Julian Togelius and Andrew Nealen",
    year = "2017",
    month = "2",
    day = "21",
    doi = "10.1109/CIG.2016.7860398",
    language = "English (US)",
    booktitle = "2016 IEEE Conference on Computational Intelligence and Games, CIG 2016",
    publisher = "IEEE Computer Society",
    address = "United States",

    }

    TY - GEN

    T1 - Hyper-heuristic general video game playing

    AU - Mendes, Andre

    AU - Togelius, Julian

    AU - Nealen, Andrew

    PY - 2017/2/21

    Y1 - 2017/2/21

    N2 - In general video game playing, the challenge is to create agents that play unseen games proficiently. Stochastic tree search algorithms, like Monte Carlo Tree Search, perform relatively well on this task. However, performance is non-transitive: different agents perform best in different games, which means that there is not a single agent that is the best in all the games. Rather, some types of games are dominated by a few agents whereas other different agents dominate other types of games. Thus, it should be possible to construct a hyper-agent that selects from a portfolio, in which constituent sub-agents will play a new game best. Since there is no knowledge about the games, the agent needs to use available features to predict the most suitable algorithm. This work constructs such a hyper-agent using the General Video Game Playing Framework (GVGAI). The proposed method achieves promising results that show the applicability of hyper-heuristics in general video game playing and related tasks.

    AB - In general video game playing, the challenge is to create agents that play unseen games proficiently. Stochastic tree search algorithms, like Monte Carlo Tree Search, perform relatively well on this task. However, performance is non-transitive: different agents perform best in different games, which means that there is not a single agent that is the best in all the games. Rather, some types of games are dominated by a few agents whereas other different agents dominate other types of games. Thus, it should be possible to construct a hyper-agent that selects from a portfolio, in which constituent sub-agents will play a new game best. Since there is no knowledge about the games, the agent needs to use available features to predict the most suitable algorithm. This work constructs such a hyper-agent using the General Video Game Playing Framework (GVGAI). The proposed method achieves promising results that show the applicability of hyper-heuristics in general video game playing and related tasks.

    UR - http://www.scopus.com/inward/record.url?scp=85015393291&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=85015393291&partnerID=8YFLogxK

    U2 - 10.1109/CIG.2016.7860398

    DO - 10.1109/CIG.2016.7860398

    M3 - Conference contribution

    AN - SCOPUS:85015393291

    BT - 2016 IEEE Conference on Computational Intelligence and Games, CIG 2016

    PB - IEEE Computer Society

    ER -