Predicting optimal solution cost with conditional probabilities

Predicting optimal solution cost

Abstract

Heuristic search algorithms are designed to return an optimal path from a start state to a goal state. They find the optimal solution cost as a side effect. However, there are applications in which all one wants to know is an estimate of the optimal solution cost. The actual path from start to goal is not initially needed. For instance, one might be interested in quickly assessing the monetary cost of a project for bidding purposes. In such cases only the cost of executing the project is required. The actual construction plan could be formulated later, after bidding. In this paper we propose an algorithm, named Solution Cost Predictor (SCP), that accurately and efficiently predicts the optimal solution cost of a problem instance without finding the actual solution. While SCP can be viewed as a heuristic function, it differs from a heuristic conceptually in that: 1) SCP is not required to be fast enough to guide search algorithms; 2) SCP is not required to be admissible; 3) our measure of effectiveness is the prediction accuracy, which is in contrast to the solution quality and number of nodes expanded used to measure the effectiveness of heuristic functions. We show empirically that SCP makes accurate predictions on several heuristic search benchmarks.

This is a preview of subscription content, access via your institution.

References

  1. 1.

    Archer, A.F.: A modern treatment of the 15-puzzle. Am. Math. Mon. 106, 793–799 (1999)

    Article  MATH  MathSciNet  Google Scholar 

  2. 2.

    Balas, E., Toth, P.: Branch and bound methods. In: Lawler, E.L., Lenstra, J.K., Rinnooy Kart, A.H.G., Shmoys, D.B. (eds.) The Traveling Salesman Problem: A Guided Tour of Combinatorial Optimization. Wiley, New York (1985)

    Google Scholar 

  3. 3.

    Burns, E., Ruml, W.: Iterative-deepening search with on-line tree size prediction. In: Proceedings of the International Conference on Learning and Intelligent Optimization, pp. 1–15 (2012)

  4. 4.

    Chen, P.-C.: Heuristic Sampling on Backtrack Trees. PhD thesis, Stanford University (1989)

  5. 5.

    Chenoweth, S.V., Davis, H.W.: High performance A* search using rapidly growing heuristics. In: International Joint Conference on Artificial Intelligence (1991)

  6. 6.

    Culberson, J.C., Schaeffer, J.: Searching with pattern databases. In: Proceedings of the Canadian Conference on Artificial Intelligence, volume 1081 of Lecture Notes in Computer Science, pp. 402–416. Springer (1996)

  7. 7.

    Dweighter, H.: Problem E2569. Am. Math. Mon. 82, 1010 (1975)

    Article  MathSciNet  Google Scholar 

  8. 8.

    Ernandes, M., Gori, M.: Likely-admissible and sub-symbolic heuristics. In: Proceedings of the European Conference on Artificial Intelligence, pp. 613–617 (2004)

  9. 9.

    Felner, A., Korf, R.E., Hanan, S.: Additive pattern database heuristics. J. Artif. Intell. Res. 22, 279–318 (2004)

    MATH  MathSciNet  Google Scholar 

  10. 10.

    Felner, A., Zahavi, U., Schaeffer, J., Holte, R.C.: Dual lookups in pattern databases. In: Proceedings of the International Joint Conference on Artificial Intelligence, pp. 103–108 (2005)

  11. 11.

    Hart, P.E., Nilsson, N.J., Raphael, B.: A formal basis for the heuristic determination of minimum cost paths. IEEE Trans. Syst. Sci. Cybern. SSC-4(2), 100–107 (1968)

    Article  Google Scholar 

  12. 12.

    Helmert, M.: Landmark heuristics for the pancake problem. In: Proceedings of the Symposium on Combinatorial Search, pp. 109–110. AAAI Press (2010)

  13. 13.

    Helmert, M., Haslum, P., Hoffmann, J.: Flexible abstraction heuristics for optimal sequential planning. In: Proceedings of the International Conference on Automated Planning and Scheduling, pp. 176–183 (2007)

  14. 14.

    Jabbari Arfaee, S., Zilles, S., Holte, R.C.: Learning heuristic functions for large state spaces. Artif. Intell. 175(16–17), 2075–2098 (2011)

    Article  MATH  MathSciNet  Google Scholar 

  15. 15.

    Kilby, P., Slaney, J.K., Thiébaux, S., Walsh, T.: Estimating search tree size. In: Proceedings of the AAAI Conference on Artificial Intelligence, pp. 1014–1019. AAAI Press (2006)

  16. 16.

    Knuth, D.E.: Estimating the efficiency of backtrack programs. Math. Comp. 29, 121–136 (1975)

    Article  MATH  MathSciNet  Google Scholar 

  17. 17.

    Korf, R.E.: Depth-first iterative-deepening: An optimal admissible tree search. Artif. Intell. 27(1), 97–109 (1985)

    Article  MATH  MathSciNet  Google Scholar 

  18. 18.

    Korf, R.E.: Linear-space best-first search. Artif. Intell. 62(1), 41–78 (1993)

    Article  MATH  MathSciNet  Google Scholar 

  19. 19.

    Korf, R.E.: Linear-time disk-based implicit graph search. J. ACM 55(6), 26:1–26:40 (2008)

    Article  MathSciNet  Google Scholar 

  20. 20.

    Korf, R.E., Felner, A.: Disjoint pattern database heuristics. Artif. Intell. 134(1–2), 9–22 (2002)

    Article  MATH  Google Scholar 

  21. 21.

    Korf, R.E., Felner, A.: Recent progress in heuristic search: A case study of the four-peg Towers of Hanoi problem. In: Proceedings of the International Joint Conference on Artificial Intelligence, pp. 2324–2329 (2007)

  22. 22.

    Korf, R.E., Reid, M., Edelkamp, S.: Time complexity of iterative-deepening-A . Artif. Intell. 129(1–2), 199–218 (2001)

    Article  MATH  MathSciNet  Google Scholar 

  23. 23.

    Lelis, L., Stern, R., Felner, A., Zilles, S., Holte, R.C.: Predicting optimal solution cost with bidirectional stratified sampling. In: Proceedings of the International Conference on Automated Planning and Scheduling, pp. 155–163. AAAI Press (2012)

  24. 24.

    Lelis, L., Stern, R., Jabbari Arfaee, S.: Predicting solution cost with condiditional probabilities. In: Proceedings of the Symposium on Combinatorial Search, pp. 100–107. AAAI Press (2011)

  25. 25.

    Lelis, L., Zilles, S., Holte, R.C.: Improved prediction of IDA*’s performance via 𝜖-truncation. In: Proceedings of the Symposium on Combinatorial Search, pp. 108–116. AAAI Press (2011)

  26. 26.

    Lelis, L.H.S.: Active stratified sampling with clustering-based type systems for predicting the search tree size of problems with real-valued heuristics. In: Proceedings of the Symposium on Combinatorial Search, pp. 123–131. AAAI Press (2013)

  27. 27.

    Lelis, L.H.S., Otten, L., Dechter, R.: Predicting the size of depth-first branch and bound search trees. In: International Joint Conference on Artificial Intelligence, pp. 594–600 (2013)

  28. 28.

    Lelis, L.H.S., Zilles, S., Holte, R.C.: Predicting the size of IDA*’s search tree. Artif. Intell., 53–76 (2013)

  29. 29.

    Pearl, J.: Probabilistic Reasoning in Intelligent Systems. Morgan Kaufmann (1988)

  30. 30.

    Prieditis, A.E.: Machine discovery of effective admissible heuristics. Mach. Learn. 12(1–3), 117–141 (1993)

    Google Scholar 

  31. 31.

    Richter, S., Helmert, M.: Preferred operators and deferred evaluation in satisficing planning. In: Proceedings of the International Conference on Automated Planning and Scheduling, pp. 273–280 (2009)

  32. 32.

    Samadi, M., Felner, A., Schaeffer, J.: Learning from multiple heuristics. In: Proceedings of the AAAI Conference on Artificial Intelligence, pp. 357–362. AAAI Press (2008)

  33. 33.

    Slocum, J., Sonneveld, D.: The 15 Puzzle. Slocum Puzzle Foundation (2006)

  34. 34.

    Stern, R., Puzis, R., Felner, A.: Potential search: a bounded-cost search algorithm. In: Proceedings of the International Conference on Automated Planning and Scheduling, pp. 234–241 (2011)

  35. 35.

    Sturtevant, N.R., Felner, A., Barrer, M., Schaeffer, J., Burch, N.: Memory-based heuristics for explicit state spaces. In: Proceedings of the International Joint Conference on Artificial Intelligence, pp. 609–614 (2009)

  36. 36.

    Thayer, J., Dionne, A., Ruml, W.: Learning inadmissible heuristics during search. In: Proceedings of the International Conference on Automated Planning and Scheduling, pp. 250–257 (2011)

  37. 37.

    Thayer, J.T., Stern, R., Lelis, L.H.S.: Are we there yet? - estimating search progress. In: Proceedings of the 5th Annual Symposium on Combinatorial Search, pp. 129–136. AAAI Press (2012)

  38. 38.

    Yang, F., Culberson, J.C., Holte, R.C., Zahavi, U., Felner, A.: A general theory of additive state space abstractions. J. Artif. Intell. Res. 32, 631–662 (2008)

    MATH  MathSciNet  Google Scholar 

  39. 39.

    Zahavi, U., Felner, A., Burch, N., Holte, R.C.: Predicting the performance of IDA* using conditional distributions. J. Artif. Intell. Res. 37, 41–83 (2010)

    MATH  MathSciNet  Google Scholar 

  40. 40.

    Zahavi, U., Felner, A., Holte, R.C., Schaeffer, J.: Duality in permutation state spaces and the dual search algorithm. Artif. Intell. 172(4–5), 514–540 (2008)

    Article  MATH  MathSciNet  Google Scholar 

Download references

Author information

Affiliations

Authors

Corresponding author

Correspondence to Levi H. S. Lelis.

Additional information

This work was carried out while L. H. S. Lelis was at the University of Alberta.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Lelis, L.H.S., Stern, R., Felner, A. et al. Predicting optimal solution cost with conditional probabilities. Ann Math Artif Intell 72, 267–295 (2014). https://doi.org/10.1007/s10472-014-9432-8

Download citation

Keywords

  • Optimal solution cost prediction
  • Type systems
  • Heuristic search

Mathematics Subject Classifications (2010)

  • 68T20
  • 68W20
  • 68W25