Learning Models of the Opponent's Strategy in Game Playing

David Carmel and Shaul Markovitch. Learning Models of the Opponent's Strategy in Game Playing. In Proceedings of The AAAI Fall Symposium on Games: Planing and Learning, 140-147 North Carolina, 1993.

Abstract

This work studies the problem of opponent modeling in game playing. A simplified version of a model is defined as a pair of search depth and evaluation function. M*, a generalization of the minimax algorithm that can handle an opponent model is described. The benefit of using opponent models is demonstrated by comparing the performance of M* with that of traditional minimax algorithm. An algorithm for learning the search depth component of the opponent's strategy is described. It is shown that the algorithm can learn even in the presence of imperfect function model. Finally, it is shown that learning the opponent's function can be done using the existing techniques for learning book games.

Keywords: Opponent Modeling, Learning in Games, Multi-Agent Systems, Games

Secondary Keywords:

Online version:

Bibtex entry:

 @inproceedings{Carmel:1993:LMO,
  Author = {David Carmel and Shaul Markovitch},
  Title = {Learning Models of the Opponent's Strategy in Game Playing},
  Year = {1993},
  Booktitle = {Proceedings of The AAAI Fall Symposium on Games: Planing and Learning},
  Pages = {140--147},
  Address = {North Carolina},
  Url = {http://www.cs.technion.ac.il/~shaulm/papers/pdf/Carmel-Markovitch-fss1993.pdf},
  Keywords = {Opponent Modeling, Learning in Games, Multi-Agent Systems, Games},
  Secondary-keywords = {Adversary Search},
  Abstract = {
    This work studies the problem of opponent modeling in game
    playing. A simplified version of a model is defined as a pair of
    search depth and evaluation function. M*, a generalization of the
    minimax algorithm that can handle an opponent model is described.
    The benefit of using opponent models is demonstrated by comparing
    the performance of M* with that of traditional minimax algorithm.
    An algorithm for learning the search depth component of the
    opponent's strategy is described. It is shown that the algorithm
    can learn even in the presence of imperfect function model.
    Finally, it is shown that learning the opponent's function can be
    done using the existing techniques for learning book games.
  }

  }