Multi-population competitive co-evolution of car racing controllers

Julian Togelius, Peter Burrow, Simon M. Lucas

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

    Abstract

    Multi-population competitive co-evolution is explored as a way of developing controllers for a simple (but definitely not trivial) car racing game. The three main uses we see for this method are to evolve more complex general intelligence than would be possible with other methods, to compare different evolvable architectures for controllers, and to develop behaviourally diverse populations of agents for computer games. Nine-population co-evolution is compared with single-population co-evolution and standard evolution strategies, steady-state and generational versions of the algorithm are compared, and a number of different controller architectures are compared with each other.

    Original languageEnglish (US)
    Title of host publication2007 IEEE Congress on Evolutionary Computation, CEC 2007
    Pages4043-4050
    Number of pages8
    DOIs
    StatePublished - 2007
    Event2007 IEEE Congress on Evolutionary Computation, CEC 2007 - , Singapore
    Duration: Sep 25 2007Sep 28 2007

    Other

    Other2007 IEEE Congress on Evolutionary Computation, CEC 2007
    CountrySingapore
    Period9/25/079/28/07

    Fingerprint

    Coevolution
    Railroad cars
    Controller
    Controllers
    Computer games
    Computer Games
    Evolution Strategies
    Trivial
    Game
    Architecture

    ASJC Scopus subject areas

    • Artificial Intelligence
    • Software
    • Theoretical Computer Science

    Cite this

    Togelius, J., Burrow, P., & Lucas, S. M. (2007). Multi-population competitive co-evolution of car racing controllers. In 2007 IEEE Congress on Evolutionary Computation, CEC 2007 (pp. 4043-4050). [4424998] https://doi.org/10.1109/CEC.2007.4424998

    Multi-population competitive co-evolution of car racing controllers. / Togelius, Julian; Burrow, Peter; Lucas, Simon M.

    2007 IEEE Congress on Evolutionary Computation, CEC 2007. 2007. p. 4043-4050 4424998.

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

    Togelius, J, Burrow, P & Lucas, SM 2007, Multi-population competitive co-evolution of car racing controllers. in 2007 IEEE Congress on Evolutionary Computation, CEC 2007., 4424998, pp. 4043-4050, 2007 IEEE Congress on Evolutionary Computation, CEC 2007, Singapore, 9/25/07. https://doi.org/10.1109/CEC.2007.4424998
    Togelius J, Burrow P, Lucas SM. Multi-population competitive co-evolution of car racing controllers. In 2007 IEEE Congress on Evolutionary Computation, CEC 2007. 2007. p. 4043-4050. 4424998 https://doi.org/10.1109/CEC.2007.4424998
    Togelius, Julian ; Burrow, Peter ; Lucas, Simon M. / Multi-population competitive co-evolution of car racing controllers. 2007 IEEE Congress on Evolutionary Computation, CEC 2007. 2007. pp. 4043-4050
    @inproceedings{7f97f76e8bef400c9bb070fdff09a269,
    title = "Multi-population competitive co-evolution of car racing controllers",
    abstract = "Multi-population competitive co-evolution is explored as a way of developing controllers for a simple (but definitely not trivial) car racing game. The three main uses we see for this method are to evolve more complex general intelligence than would be possible with other methods, to compare different evolvable architectures for controllers, and to develop behaviourally diverse populations of agents for computer games. Nine-population co-evolution is compared with single-population co-evolution and standard evolution strategies, steady-state and generational versions of the algorithm are compared, and a number of different controller architectures are compared with each other.",
    author = "Julian Togelius and Peter Burrow and Lucas, {Simon M.}",
    year = "2007",
    doi = "10.1109/CEC.2007.4424998",
    language = "English (US)",
    isbn = "1424413400",
    pages = "4043--4050",
    booktitle = "2007 IEEE Congress on Evolutionary Computation, CEC 2007",

    }

    TY - GEN

    T1 - Multi-population competitive co-evolution of car racing controllers

    AU - Togelius, Julian

    AU - Burrow, Peter

    AU - Lucas, Simon M.

    PY - 2007

    Y1 - 2007

    N2 - Multi-population competitive co-evolution is explored as a way of developing controllers for a simple (but definitely not trivial) car racing game. The three main uses we see for this method are to evolve more complex general intelligence than would be possible with other methods, to compare different evolvable architectures for controllers, and to develop behaviourally diverse populations of agents for computer games. Nine-population co-evolution is compared with single-population co-evolution and standard evolution strategies, steady-state and generational versions of the algorithm are compared, and a number of different controller architectures are compared with each other.

    AB - Multi-population competitive co-evolution is explored as a way of developing controllers for a simple (but definitely not trivial) car racing game. The three main uses we see for this method are to evolve more complex general intelligence than would be possible with other methods, to compare different evolvable architectures for controllers, and to develop behaviourally diverse populations of agents for computer games. Nine-population co-evolution is compared with single-population co-evolution and standard evolution strategies, steady-state and generational versions of the algorithm are compared, and a number of different controller architectures are compared with each other.

    UR - http://www.scopus.com/inward/record.url?scp=54849154600&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=54849154600&partnerID=8YFLogxK

    U2 - 10.1109/CEC.2007.4424998

    DO - 10.1109/CEC.2007.4424998

    M3 - Conference contribution

    SN - 1424413400

    SN - 9781424413409

    SP - 4043

    EP - 4050

    BT - 2007 IEEE Congress on Evolutionary Computation, CEC 2007

    ER -