Learning Effective Search Heuristics

SAGE.2 is a production system that improves its search strategies with practice. The program incorporates four different heuristics for assigning credit and blame, and employs a discrimination process to direct its search through the space of move-proposing rules. The system has shown its generality by learning search heuristics in five different task domains. In addition to improving its search behavior on practice problems, SAGE.2 was able to transfer its expertise to scaled-up versions of a task, and in one case transferred its acquired search strategy to problems with different initial and goal states.