Automated Discovery of Search-Extension Features
One of the main challenges with selective search extensions is designing effective move categories (features). Usually, it is a manual trial-and-error task, which requires both intuition and expert human knowledge. Automating this task potentially enables the discovery of both more complex and more effective move categories. The current work introduces Gradual Focus, an algorithm for automatically discovering interesting move categories for selective search extensions. The algorithm iteratively creates new more refined move categories by combining features from an atomic feature set. Empirical data is presented for the game Breakthrough showing that Gradual Focus looks at a number of combinations that is two orders of magnitude fewer than a brute-force method does, while preserving adequate precision and recall.
KeywordsAtomic Feature Pruning Method Move Category Interesting Combination Neutral Feature
Unable to display preview. Download preview PDF.
- 2.Beal, D.F., Smith, M.C.: Quantification of search extension benefits. ICCA Journal 8(4), 205–218 (1995)Google Scholar
- 3.Hyatt, R.M.: Crafty. A chess program (1996) (March 27, 2008), ftp://ftp.cis.uab.edu/pub/hyatt
- 4.Levy, D., Broughton, D., Taylor, M.: The SEX algorithm in computer chess. ICCA Journal 12(1), 10–21 (1989)Google Scholar
- 5.Tsuruoka, Y., Yokoyama, D., Chikayama, T.: Game-tree search algorithm based on realization probability. ICGA Journal 25(3), 146–153 (2002)Google Scholar
- 7.Björnsson, Y.: Selective Depth-First Game-Tree Search. Phd dissertation, University of Alberta (2002)Google Scholar
- 10.Fawcett, T.E., Utgoff, P.E.: Automatic feature generation for problem solving systems. In: Intern. Conf. on Machine Learning (ICML), pp. 144–153 (1992)Google Scholar
- 11.Kaneko, T., Yamaguchi, K., Kawai, S.: Automated identification of patterns in evaluation functions. In: Advances in Computer Games, vol. 10, pp. 279–298 (2003)Google Scholar
- 13.Buro, M.: Experiments with Multi-ProbCut and a new high-quality evaluation function for Othello. In: Games in AI Research, pp. 77–96 (1999)Google Scholar
- 15.Finkelstein, L., Markovitch, S.: Learning to play chess selectively by acquiring move patterns. ICCA Journal 21(2), 100–119 (1998)Google Scholar
- 16.Handscomb, K.: 8×8 game design competition: The winning game: Breakthrough ... and two other favorites. Abstract Games Magazine 7 (2001)Google Scholar