Online evolution for multi-action adversarial games

Niels Justesen, Tobias Mahlmann, Julian Togelius

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

    Abstract

    We present Online Evolution, a novel method for playing turn-based multi-action adversarial games. Such games, which include most strategy games, have extremely high branching factors due to each turn having multiple actions. In Online Evolution, an evolutionary algorithm is used to evolve the combination of atomic actions that make up a single move, with a state evaluation function used for fitness. We implement Online Evolution for the turn-based multi-action game Hero Academy and compare it with a standard Monte Carlo Tree Search implementation as well as two types of greedy algorithms. Online Evolution is shown to outperform these methods by a large margin. This shows that evolutionary planning on the level of a single move can be very effective for this sort of problems.

    Original languageEnglish (US)
    Title of host publicationApplications of Evolutionary Computation - 19th European Conference, EvoApplications 2016, Proceedings
    PublisherSpringer Verlag
    Pages590-603
    Number of pages14
    Volume9597
    ISBN (Print)9783319312033
    DOIs
    StatePublished - 2016
    Event19th European Conference on Applications of Evolutionary Computation, EvoApplications 2016 - Porto, Portugal
    Duration: Mar 30 2016Apr 1 2016

    Publication series

    NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    Volume9597
    ISSN (Print)03029743
    ISSN (Electronic)16113349

    Other

    Other19th European Conference on Applications of Evolutionary Computation, EvoApplications 2016
    CountryPortugal
    CityPorto
    Period3/30/164/1/16

    Fingerprint

    Function evaluation
    Evolutionary algorithms
    Game
    Planning
    Search Trees
    Evaluation Function
    Greedy Algorithm
    Margin
    Sort
    Fitness
    Branching
    Evolutionary Algorithms

    ASJC Scopus subject areas

    • Computer Science(all)
    • Theoretical Computer Science

    Cite this

    Justesen, N., Mahlmann, T., & Togelius, J. (2016). Online evolution for multi-action adversarial games. In Applications of Evolutionary Computation - 19th European Conference, EvoApplications 2016, Proceedings (Vol. 9597, pp. 590-603). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9597). Springer Verlag. https://doi.org/10.1007/978-3-319-31204-0_38

    Online evolution for multi-action adversarial games. / Justesen, Niels; Mahlmann, Tobias; Togelius, Julian.

    Applications of Evolutionary Computation - 19th European Conference, EvoApplications 2016, Proceedings. Vol. 9597 Springer Verlag, 2016. p. 590-603 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9597).

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

    Justesen, N, Mahlmann, T & Togelius, J 2016, Online evolution for multi-action adversarial games. in Applications of Evolutionary Computation - 19th European Conference, EvoApplications 2016, Proceedings. vol. 9597, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9597, Springer Verlag, pp. 590-603, 19th European Conference on Applications of Evolutionary Computation, EvoApplications 2016, Porto, Portugal, 3/30/16. https://doi.org/10.1007/978-3-319-31204-0_38
    Justesen N, Mahlmann T, Togelius J. Online evolution for multi-action adversarial games. In Applications of Evolutionary Computation - 19th European Conference, EvoApplications 2016, Proceedings. Vol. 9597. Springer Verlag. 2016. p. 590-603. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-31204-0_38
    Justesen, Niels ; Mahlmann, Tobias ; Togelius, Julian. / Online evolution for multi-action adversarial games. Applications of Evolutionary Computation - 19th European Conference, EvoApplications 2016, Proceedings. Vol. 9597 Springer Verlag, 2016. pp. 590-603 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
    @inproceedings{5c6c4c7e9c0241559e729c14a33d5a33,
    title = "Online evolution for multi-action adversarial games",
    abstract = "We present Online Evolution, a novel method for playing turn-based multi-action adversarial games. Such games, which include most strategy games, have extremely high branching factors due to each turn having multiple actions. In Online Evolution, an evolutionary algorithm is used to evolve the combination of atomic actions that make up a single move, with a state evaluation function used for fitness. We implement Online Evolution for the turn-based multi-action game Hero Academy and compare it with a standard Monte Carlo Tree Search implementation as well as two types of greedy algorithms. Online Evolution is shown to outperform these methods by a large margin. This shows that evolutionary planning on the level of a single move can be very effective for this sort of problems.",
    author = "Niels Justesen and Tobias Mahlmann and Julian Togelius",
    year = "2016",
    doi = "10.1007/978-3-319-31204-0_38",
    language = "English (US)",
    isbn = "9783319312033",
    volume = "9597",
    series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
    publisher = "Springer Verlag",
    pages = "590--603",
    booktitle = "Applications of Evolutionary Computation - 19th European Conference, EvoApplications 2016, Proceedings",

    }

    TY - GEN

    T1 - Online evolution for multi-action adversarial games

    AU - Justesen, Niels

    AU - Mahlmann, Tobias

    AU - Togelius, Julian

    PY - 2016

    Y1 - 2016

    N2 - We present Online Evolution, a novel method for playing turn-based multi-action adversarial games. Such games, which include most strategy games, have extremely high branching factors due to each turn having multiple actions. In Online Evolution, an evolutionary algorithm is used to evolve the combination of atomic actions that make up a single move, with a state evaluation function used for fitness. We implement Online Evolution for the turn-based multi-action game Hero Academy and compare it with a standard Monte Carlo Tree Search implementation as well as two types of greedy algorithms. Online Evolution is shown to outperform these methods by a large margin. This shows that evolutionary planning on the level of a single move can be very effective for this sort of problems.

    AB - We present Online Evolution, a novel method for playing turn-based multi-action adversarial games. Such games, which include most strategy games, have extremely high branching factors due to each turn having multiple actions. In Online Evolution, an evolutionary algorithm is used to evolve the combination of atomic actions that make up a single move, with a state evaluation function used for fitness. We implement Online Evolution for the turn-based multi-action game Hero Academy and compare it with a standard Monte Carlo Tree Search implementation as well as two types of greedy algorithms. Online Evolution is shown to outperform these methods by a large margin. This shows that evolutionary planning on the level of a single move can be very effective for this sort of problems.

    UR - http://www.scopus.com/inward/record.url?scp=84961750938&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84961750938&partnerID=8YFLogxK

    U2 - 10.1007/978-3-319-31204-0_38

    DO - 10.1007/978-3-319-31204-0_38

    M3 - Conference contribution

    SN - 9783319312033

    VL - 9597

    T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

    SP - 590

    EP - 603

    BT - Applications of Evolutionary Computation - 19th European Conference, EvoApplications 2016, Proceedings

    PB - Springer Verlag

    ER -