Extension Heuristics¹

Article type: Research Article

Authors: Anantharaman, T.S.

Affiliations: Tudor Investment Corporation, Greenwich, CT

Note: [1] This article is an edited version of Chapter 4 of the author’s Ph.D. thesis, May 1990, Carnegie-Mellon University (Report CMU-CS-90-173). The editing was mainly done in order to make the contribution self-contained and no information has been altered or omitted. This article elaborates the research on statistical measures as reported in “Confidently Selecting a Search Heuristic” in the ICCA Journal, Vol. 14, No. 1, pp. 3–16.

Abstract: This article lists a number of extension heuristics which will be evaluated in terms of the performance metric described in Anantharaman (1991), specifically that of line 11 in Table 2. The heuristics have all been implemented in the Deep Thought program used for the experiments, but not all are present in the (older) version of Deep Thought that was used to earn a USCF rating of 2551 in tournament games against human players. After describing each heuristic algorithmically, some implementation details are provided where appropriate, and finally the experimental result for the heuristic is discussed. All experimental results are tabulated in an Appendix.

DOI: 10.3233/ICG-1991-14202

Journal: ICGA Journal, vol. 14, no. 2, pp. 47-65, 1991