Informed Search Using Equivalent-Class Templates

Many of the search techniques used in AI were developed, studied and/or optimized in the context of game playing. Game playing demands efficient methods and has the advantage that the testing and evaluation of different implementations is straightforward, for it requires playing the game. This paper provides a case study of an informed search technique applied for a non-adversarial game (triangular board Peg Game). The performance of the method is compared with other strategies that were tried. In addition, conclusions about the peg game are presented.