Document Open Access Logo

A Portfolio-Based Approach to Select Efficient Variable Ordering Heuristics for Constraint Satisfaction Problems

Authors Hongbo Li , Yaling Wu, Minghao Yin, Zhanshan Li

PDF
Thumbnail PDF

File

LIPIcs.CP.2022.32.pdf
  • Filesize: 0.73 MB
  • 10 pages

Document Identifiers

Author Details

Hongbo Li
  • School of Information Science and Technology, Northeast Normal University, Changchun, China
Yaling Wu
  • School of Information Science and Technology, Northeast Normal University, Changchun, China
Minghao Yin
  • School of Information Science and Technology, Northeast Normal University, Changchun, China
Zhanshan Li
  • College of Computer Science and Technology, Jilin University, Changchun, China

Funding

  • Li, Hongbo: The National Natural Science Foundation of China (61802056). The Plan of Science and Technology Development in Jilin Province (20210101470JC). Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education, Jilin University.
  • Yin, Minghao: The National Natural Science Foundation of China under Grant NO. 61976050.
  • Li, Zhanshan: Open Research Fund of Key Laboratory of Space Utilization, Chinese Academy of Sciences under Grant NO. LSU-KFJJ-2019-08.

Cite AsGet BibTex

Hongbo Li, Yaling Wu, Minghao Yin, and Zhanshan Li. A Portfolio-Based Approach to Select Efficient Variable Ordering Heuristics for Constraint Satisfaction Problems. In 28th International Conference on Principles and Practice of Constraint Programming (CP 2022). Leibniz International Proceedings in Informatics (LIPIcs), Volume 235, pp. 32:1-32:10, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2022)
https://doi.org/10.4230/LIPIcs.CP.2022.32

Abstract

Variable ordering heuristics (VOH) play a central role in solving Constraint Satisfaction Problems (CSP). The performance of different VOHs may vary greatly in solving the same CSP instance. In this paper, we propose an approach to select efficient VOHs for solving different CSP instances. The approach contains two phases. The first phase is a probing procedure that runs a simple portfolio strategy containing several different VOHs. The portfolio tries to use each of the candidate VOHs to guide backtracking search to solve the CSP instance within a limited number of failures. If the CSP is not solved by the portfolio, one of the candidates is selected for it by analysing the information attached in the search trees generated by the candidates. The second phase uses the selected VOH to guide backtracking search to solve the CSP. The experiments are run with the MiniZinc benchmark suite and four different VOHs which are considered as the state of the art are involved. The results show that the proposed approach finds the best VOH for more than 67% instances and it solves more instances than all the candidate VOHs and an adaptive VOH based on Multi-Armed Bandit. It could be an effective adaptive search strategy for black-box CSP solvers.

Subject Classification

ACM Subject Classification
  • Theory of computation → Constraint and logic programming
Keywords
  • Constraint Satisfaction Problem
  • Variable Ordering Heuristic
  • Adaptive Search Heuristic
  • Portfolio

Metrics

  • Access Statistics
  • Total Accesses (updated on a weekly basis)
    0
    PDF Downloads
    0
    Metadata Views

Supplementary Materials

References

  1. R. Amadini, M. Gabbrielli, and J. Mauro. Portfolio approaches for constraint optimization problems. Annals of Mathematics and Artificial Intelligence, 76:229-246, 2016. Google Scholar
  2. A. Balafrej, C. Bessiere, and A. Paparrizou. Multi-armed bandits for adaptive constraint propagation. In Proc. IJCAI'15, pages 290-296. AAAI Press, 2015. Google Scholar
  3. F. Boussemart, F. Hemery, C. Lecoutre, and L. Sais. Boosting systematic search by weighting constraints. In Proc. ECAI'04, pages 146-150, 2004. Google Scholar
  4. G. Chu and P. J. Stuckey. Learning value heuristics for constraint programming. In Proc. CPAIOR'15, pages 108-123. Springer, 2015. Google Scholar
  5. S. Epstein and S. Petrovic. Learning to solve constraint problems. In Proc. ICAPS'07, Workshop on Planning and Learning, 2007. Google Scholar
  6. C. P. Gomes, B. Selman, and H. Kautz. Boosting combinatorial search through randomization. In Proc. AAAI'98, pages 431-437. AAAI, 1998. Google Scholar
  7. D. Habet and C. Terrioux. Conflict history based search for constraint satisfaction problem. In Proc. of the 34th ACM/SIGAPP Symposium on Applied Computing, pages 1117-1122. ACM, 2019. Google Scholar
  8. R. Haralick and G. Elliott. Increasing tree search efficiency for constraint satisfaction problems. Artificial Intelligence, 14:263-313, 1980. Google Scholar
  9. H. Hurley, L. Kotthoff, Y. Malitsky, and B. O’Sullivan. Proteus: a hierarchical portfolio of solvers and transformations. In Proc. CPAIOR'14, 2014. Google Scholar
  10. J. Hwang and D. G. Mitchell. 2-way vs d-way branching for csp. In Proc. CP'05, pages 343-357. Springer, 2005. Google Scholar
  11. H. Li, J. H. Lee, H. Mi, and M. Yin. Finding good subtrees for constraint optimization problems using frequent pattern mining. In Proc. AAAI'20, pages 1577-1584, 2020. Google Scholar
  12. H. Li, Y. Liang, N. Zhang, J. Guo, D. Xu, and Z. Li. Improving degree-based variable ordering heuristics for solving constraint satisfaction problems. Journal of Heuristics, 22(2):125-145, 2016. Google Scholar
  13. H. Li, M. Yin, and Z. Li. Failure Based Variable Ordering Heuristics for Solving CSPs. In Proc. CP'21, pages 9:1-9:10, 2021. Google Scholar
  14. P. Liberatore. On the complexity of choosing the branching literal in dpll. Artificial Intelligence, 116(1):315-326, 2000. Google Scholar
  15. M. Loth, M. Sebag, Y. Hamadi, and M. Schoenauer. Bandit-based search for constraint programming. In Proc. CP'13, pages 464-480. Springer, 2013. Google Scholar
  16. L. Michel and P. Van Hentenryck. Activity-based search for black-box constraint programming solvers. In Proc. CPAIOR'12, pages 228-243. Springer, 2012. Google Scholar
  17. A. Palmieri and G. Perez. Objective as a feature for robust search strategies. In Proc. CP'18, pages 328-344. Springer, 2018. Google Scholar
  18. Anastasia Paparrizou and Kostas Stergiou. Evaluating simple fully automated heuristics for adaptive constraint propagation. In Proc. of ICTAI'12, volume 1, pages 880-885, 2012. URL: https://doi.org/10.1109/ICTAI.2012.123.
  19. G. Pesant, C. G. Quimper, and A. Zanarini. Counting-based search: Branching heuristics for constraint satisfaction problems. Journal of Artificial Intelligence Research, 43:173-210, 2012. Google Scholar
  20. C. Prud'homme, J-G. Fages, and X. Lorca. Choco Documentation. TASC - LS2N CNRS UMR 6241, COSLING S.A.S., 2017. URL: http://www.choco-solver.org.
  21. P. Refalo. Impact-based search strategies for constraint programming. In Proc. CP'04, pages 557-571. Springer, 2004. Google Scholar
  22. T. Walsh. Search in a small world. In Proc. IJCAI'99, pages 1172-1177, 1999. Google Scholar
  23. R. Wang, W. Xia, and R. H. C. Yap. Correlation heuristics for constraint programming. In Proc. ICTAI'17, pages 1037-1041. IEEE, 2017. Google Scholar
  24. H. Wattez, F. Koriche, C. Lecoutre, A. Paparrizou, and S. Tabary. Learning variable ordering heuristics with multi-armed bandits and restarts. In Proc. ECAI'20, pages 371-378. IOS Press, 2020. Google Scholar
  25. H. Wattez, C. Lecoutre, A. Paparrizou, and S. Tabary. Refining constraint weighting. In Proc. of ICTAI'19, pages 71-77. IEEE, 2019. Google Scholar
  26. R. J. Woodward, A. Schneider, B.Y. Choueiry, and C. Bessiere. Adaptive parameterized consistency for non-binary csps by counting supports. In Proc. of CP'14, pages 755-764, 2014. Google Scholar
  27. W. Xia and R. H. C. Yap. Learning robust search strategies using a bandit-based approach. In Proc. AAAI'18, pages 6657-6665. AAAI, 2018. Google Scholar
  28. L. Xu, F. Hutter, H. H Hoos, and K. Leyton-Brown. SATzilla: portfolio-based algorithm selection for SAT. Journal of Artificial Intelligence Research, 32:565-606, 2008. Google Scholar
Questions / Remarks / Feedback
X

Feedback for Dagstuhl Publishing


Thanks for your feedback!

Feedback submitted

Could not send message

Please try again later or send an E-mail
© 2023-2024 Schloss Dagstuhl – LZI GmbH Imprint Privacy Contact