Statistical Selection Among Problem-Solving Methods.

Abstract : The choice of an appropriate problem solving method, from available methods, is a crucial skill for human experts in many areas. We describe a technique for automatic selection among methods, based on a statistical analysis of their past performances. We formalize the statistical problem involved in selecting an efficient problem solving method, derive a solution to this problem, and describe a selection algorithm. The algorithm not only chooses among available methods, but also decides when to abandon the chosen method, if it proves to take too much time. We extend our basic statistical technique to account for problem sizes and for similarity between problems. We give empirical results of the use of this technique to select among search engines in the PRODIGY system. We also test the selection technique on artificially generated performance data, using several different probability distributions.

[1]  Eric Horvitz,et al.  Reasoning under Varying and Uncertain Resource Constraints , 1988, AAAI.

[2]  Eric Horvitz,et al.  Ideal reformulation of belief networks , 1990, UAI.

[3]  Herbert A. Simon,et al.  CaMeRa: A Computational Model of Multiple Representations , 1997, Cogn. Sci..

[4]  Shlomo Zilberstein,et al.  Monitoring the Progress of Anytime Problem-Solving , 1996, AAAI/IAAI, Vol. 2.

[5]  Manuela M. Veloso,et al.  The Need for Different Domain-independent Heuristics , 1994, AIPS.

[6]  William W. Cohen USING DISTRIBUTION‐FREE LEARNING THEORY TO ANALYZE SOLUTION‐PATH CACHING MECHANISMS , 1992, Comput. Intell..

[7]  Stuart J. Russell Fine-Grained Decision-Theoretic Search Control , 2013, UAI 1990.

[8]  William Mendenhall,et al.  Introduction to Probability and Statistics , 1968 .

[9]  Allen Newell,et al.  Human Problem Solving. , 1973 .

[10]  Steven Minton,et al.  Selecting the Right Heuristic Algorithm: Runtime Performance Predictors , 1996, Canadian Conference on AI.

[11]  Craig A. Knoblock Generating abstraction hierarchies - an automated approach to reducing search in planning , 1993, The Kluwer international series in engineering and computer science.

[12]  Manuela M. Veloso,et al.  Planning and Learning by Analogical Reasoning , 1994, Lecture Notes in Computer Science.

[13]  Manuela M. Veloso,et al.  FLECS: Planning with a Flexible Commitment Strategy , 1995, J. Artif. Intell. Res..

[14]  Eugene Fink,et al.  Automatic representation changes in problem solving , 1999 .

[15]  Eugene Fink,et al.  Integrating planning and learning: the PRODIGY architecture , 1995, J. Exp. Theor. Artif. Intell..

[16]  Michael Patkin Models of Thought, Volume II by Herbert Simon (Yale University Press, New Haven and London, 1989) pp.xviii + 508, ISBN 0-300-04230-2 , 1990 .

[17]  Maria Alicia Perez Learning search control knowledge to improve plan quality , 1996 .

[18]  Craig A. Knoblock Automatically generating abstractions for problem solving , 1991 .

[19]  David Haussler,et al.  Occam's Razor , 1987, Inf. Process. Lett..

[20]  Qiang Yang,et al.  The Expected Value of Hierarchical Problem-Solving , 1992, AAAI.

[21]  Devika Subramanian,et al.  Provably Bounded Optimal Agents , 1993, IJCAI.

[22]  Adele E. Howe,et al.  Exploiting Competitive Planner Performance , 1999, ECP.

[23]  S.J.J. Smith,et al.  Empirical Methods for Artificial Intelligence , 1995 .

[24]  Craig A. Knoblock Automatically Generating Abstractions for Planning , 1994, Artif. Intell..

[25]  Leslie G. Valiant,et al.  A theory of the learnable , 1984, STOC '84.