Analysis of Heuristic Synergies
“Heuristic synergy” refers to improvements in search performance when the decisions made by two or more heuristics are combined. This paper considers combinations based on products and quotients, and a less familiar form of combination based on weighted sums of ratings from a set of base heuristics, some of which result in definite improvements in performance. Then, using recent results from a factor analytic study of heuristic performance, which had demonstrated two main effects of heuristics involving either buildup of contention or look-ahead-induced failure, it is shown that heuristic combinations are effective when they are able to balance these two actions. In addition to elucidating the basis for heuristic synergy (or lack thereof), this work suggests that the task of understanding heuristic search depends on the analysis of these two basic actions.
KeywordsRatio Node Constraint Satisfaction Problem Search Node Heuristic Performance Improve Search Performance
Unable to display preview. Download preview PDF.
- 1.Bessière, C., Régin, J.-C.: Mac and combined heuristics: Two reasons to forsake fc (and cbj?) on hard problems. In: Freuder, E.C. (ed.) CP 1996. LNCS, vol. 1118, pp. 61–75. Springer, Heidelberg (1996)Google Scholar
- 6.Smith, B.M., Grant, S.A.: Trying harder to fail first. In: Proc. Thirteenth European Conference on Artificial Intelligence-ECAI 1998, pp. 249–253. John Wiley & Sons, Chichester (1998)Google Scholar
- 7.Geelen, P.A.: Dual viewpoint heuristics for binary constraint satisfaction problems. In: Proc. Tenth European Conference on Artificial Intelligence-ECAI 1992, pp. 31–35 (1992)Google Scholar
- 8.Wallace, R.J.: Csp heuristics categorized with factor analytic. In: Creaney, N. (ed.) Proc. Sixteenth Irish Conference on Artificial Intelligence and Cognitive Science, Coleraine, NI, University of Ulster, pp. 213–222 (2005)Google Scholar
- 11.Bessière, C., Zanuttini, B., Fernández, C.: Measuring search trees. In: ECAI 2004 Workshop on Modelling and Solving Problems with Constraints, pp. 31–40 (2004)Google Scholar
- 12.Boussemart, F., Hemery, F., Lecoutre, C., Sais, L.: Boosting systematic search by weighting constraints. In: Proc. Sixteenth European Conference on Artificial Intelligence-ECAI 2004, pp. 146–150 (2004)Google Scholar
- 13.Gent, I., MacIntyre, E., Prosser, P., Smith, B., Walsh, T.: An empirical study of dynamic variable ordering heuristics for the constraint satisfaction problem. In: Freuder, E.C. (ed.) CP 1996. LNCS, vol. 1118, pp. 179–193. Springer, Heidelberg (1996)Google Scholar