Welcome to the CS 110 Home Page

Advanced AI Seminar
Artificial Opponents: Machine Learning in Competitive Environments

Prof. Jordan Pollack



Sevan G. Ficici, Hugues Juillé, TAs.


Meeting Times: Monday, Wednesday, 2-3:30pm

This reading seminar covers machine learning, evolutionary, and neural computational approaches to games and other competitive environments. This is broadly defined to include classic board games like TTT, Checkers, Chess and Backgammon, abstract mathematical games, such as the prisoner's dilemma and pursuer-evader differential games, artificial life situations, and even video-like games.

Topics and readings include: Basic Game Theory, touching on classic matrix games in biological, social and economic and military systems, the basic AI programming approaches to board games, as well as machine learning, neural computational, evolutionary and co-evolutionary computation in the setting of both abstract and creative competitive environments. Student work involves group presentation and written critical revies of material from the literature, and a term project.


NEWSFLASHS

  • April 24th class has been cancelled!
  • Presentations of Projects in Progress are on April 17th, 22nd, and 29th. Plan 10-15 minutes.
  • Final projects reports, 10 pages double spaced, with references, are due in class May 1st.



  • Current Presentation Schedule

    Date Topic Presented By Reviewed By
    2/5 Axelrod Ch. 1-3, 8 Ficici Raymond
    2/7 Axelrod GA paper Shubina
    Rozier
    Helfgott
    Spitz
    2/12 Lindgren Melnick
    Funes
    Raisbeck
    Zhao
    2/14 Maynard Smith/Axelrod Ch. 5 Kitts
    Helfgott
    Raymond
    Sanders
    2/26 Isaacs, Differential Games Raisbeck
    Peng
    Zhao
    Helfgott
    Funes
    2/28 Miller and Cliff, PE Game Werger
    Sanders
    Sanchez
    3/4 Reynolds, PE Tag Ficici
    Kozubal
    Kontoglis
    Shubina
    Raisbeck
    3/6 Koza, PE/Pacman Sanchez
    Stux
    Funes
    Spitz
    3/11 Pengi & Rogomatic Shubina
    Sanders
    Laurence
    Werger
    Ficici
    Kontoglis
    3/13 Bolo/Bolo Brains Raymond
    Kozubal
    Stux
    Peng
    Laurence
    3/18 Samuels 1, Checkers Rozier
    Stux
    Zhao
    Sanchez
    Kozubal
    3/20 Samuels 2, Checkers Werger
    Spitz
    Kontoglis
    Shubina
    Peng
    3/25 Epstein, HOYLE Helfgott
    Blair
    Rozier
    Sanders
    3/27 Berliner, BKG Laurence
    Peng
    Ficici
    Kontoglis
    4/1 Tesauro, TD-Gammon Raisbeck
    Funes
    Laurence
    Zhao
    4/2 3:30 Pollack, HC-Gammon Juille
    Spitz
    Rozier
    Stux
    4/15 Sims, Robot Game Raymond
    Sanchez
    Werger
    Kozubal


    Back to top.


    Project Presentation Schedule

    Date Presented By
    4/17 Shubina
    Rozier
    Ficici
    4/22 Helfgott
    Funes
    Kontaglis
    Spitz
    Raymond
    Peng
    4/29 Kozabal
    Sanders
    Zhao
    Sanchez
    Lawrence
    Werger


    Back to top.

    Student Projects


    Back to top.

    Bibliography

    Agre, P. and Chapman, D. (1987). Pengi: An implementation of a theory of activity. In Proceedings National Conference on AI, pages 268-272.

    Axelrod, R. (1984). The evolution of cooperation. Basic Books, New York.

    Axelrod, R. (1987). The evolution of strategies in the iterated prisoner's dilemma. In Davis, L., editor, Genetic Algorithms and Simulated Annealing. Pitman: London.

    Berliner, H.J. (1980). Backgammon program beats world champion. In Artificial Intelligence, 14:205-220.

    Epstein, S.L. (1994). Toward an ideal trainer. Machine Learning, 15(3).

    Issacs, R. (1965). Differential Games. John Wiley & Sons.

    Koza, J. (1992). Genetic Programming. MIT Press, Cambridge.

    Lindgren, K. (1992). Evolutionary phenomena in simple dynamics. In C.Langton, C.Taylor, J.F. and Rasmussen, S., editors, Artificial Life II. Addison-Wesley, Reading, MA.

    Mauldin, M.L., Jacobson, G.J., Appel, A.W., and Hamey, L. G.C. (1984). Rogomatic: A belligerent expert system. In Proceedings of the Fifth National Conference of Canadian Society for Computational Studies of Intelligence.

    Miller, G.F. and Cliff, D. (1994). Protean behavior in dynamic games. In D.Cliff, P. Husbands, J. Meyer. S.W., editor, From Animals to Animats 3: Proceedings of the Third International Conference on Simulation of Adaptive Behavior.

    Pollack, J., Blair, A., and Land, M. (1995). Coevolution of a backgammon player. Technical report.

    Reynolds, Craig W. (1994). Competition, coevolution, and the game of tag. In Proceedings of the Fourth Artificial Life Conference. MIT Press.

    Samuel, A.L. (1959). Some studies of machine learning using the game of checkers. IBM Joural of Research and Development.

    Samuel, A.L. (1967). Some studies in machine learning using the game of checkers. II - recent progress. IBM Journal of Research & Development, 11(4):601-618.

    Sims, Karl (1994). Evolving 3d morphology and behavior by competition. In Proceedings of the fourth Artificial Life Conference. MIT Press.

    Smith, J.M. (1982). Evolution and the Theory of Games. Cambridge University Press.

    Tesauro, G. (1992). Practical issues in temporal difference learning. Machine Learning, 8:257-277.


    Back to top.

    Interesting Links

    Jay Scott from Swarthmore has been collecting URL's on machine learning in games.

    Stephen Intille was the source for some material on Bolo.

    There are free GP simulators (in C) available on the net, like Lil-gp and SGPC, which is in the GA-LIST library.

    Professor Roger McCain of Drexel University has some nice notes on Game Theory

    Anything one wants to know about video games begins here.


    Back to top.