Abstract
Decomposition is a well-known algorithmic technique for Graphical Models. It is commonly believed that such a technique is cost-effective for instances with low width. In this paper, we show on a large data set of real-life inspired instances that this is not the case. To better understand this result, we narrow our study and consider k-tree instances where the width is well controlled and get similar results. Finally, we show that by adding a few simple features and using simple Machine Learning models we can predict the convenience to decompose with an accuracy of more than \(85\%\), which produces time reductions in standard benchmarks of nearly \(90\%\).
Supported by grant PID2021-122830OB-C43, funded by MCIN/AEI/10.13039/501100011033 and by “ERDF: A way of making Europe”.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
A line of work under the name of soft global constraints extends this definition by allowing large scope cost functions as long as the functions are tractable.
- 2.
Note that this corresponds to solving the problem with dynamic programming implemented with memoization.
- 3.
- 4.
- 5.
References
Allen, D., Darwiche, A.: New advances in inference by recursive conditioning. In: Meek, C., Kjærulff, U. (eds.) UAI 2003, Acapulco, Mexico, 7–10 August 2003, pp. 2–10. Morgan Kaufmann (2003)
Allouche, D., de Givry, S., Katsirelos, G., Schiex, T., Zytnicki, M.: Anytime hybrid best-first search with tree decomposition for weighted CSP. In: Pesant, G. (ed.) CP 2015. LNCS, vol. 9255, pp. 12–29. Springer, Cham (2015). https://doi.org/10.1007/978-3-319-23219-5_2
Arbelaez, A., Hamadi, Y., Sebag, M.: Continuous search in constraint programming. In: Hamadi, Y., Monfroy, E., Saubion, F. (eds.) Autonomous Search, pp. 219–243. Springer, Heidelberg (2011). https://doi.org/10.1007/978-3-642-21434-9_9
Bello, I., Pham, H., Le, Q.V., Norouzi, M., Bengio, S.: Neural combinatorial optimization with reinforcement learning. In: ICLR 2017, Toulon, France, 24–26 April 2017, Workshop Track Proceedings (2017)
Bertelè, U., Brioschi, F.: On non-serial dynamic programming. J. Comb. Theory Ser. A 14(2), 137–148 (1973)
Bonfietti, A., Lombardi, M., Milano, M.: Embedding decision trees and random forests in constraint programming. In: Michel, L. (ed.) CPAIOR 2015. LNCS, vol. 9075, pp. 74–90. Springer, Cham (2015). https://doi.org/10.1007/978-3-319-18008-3_6
Cabon, B., de Givry, S., Lobjois, L., Schiex, T., Warners, J.P.: Radio link frequency assignment. Constraints Int. J. 4(1), 79–89 (1999). https://doi.org/10.1023/A:1009812409930
Cappart, Q., Moisan, T., Rousseau, L.M., Prémont-Schwarz, I., Cire, A.A.: Combining reinforcement learning and constraint programming for combinatorial optimization. In: Proceedings of the AAAI Conference on Artificial Intelligence, vol. 35, pp. 3677–3687 (2021)
Cooper, G.F.: Bayesian belief-network inference using recursive decomposition. Technical report, Knowledge Systems Laboratory, Stanford, CA (1990)
Cooper, M.C., de Givry, S., Sánchez-Fibla, M., Schiex, T., Zytnicki, M., Werner, T.: Soft arc consistency revisited. Artif. Intell. 174(7–8), 449–478 (2010)
Darwiche, A.: Recursive conditioning. Artif. Intell. 126(1–2), 5–41 (2001)
Darwiche, A.: Modeling and Reasoning with Bayesian Networks. Cambridge University Press, Cambridge (2009)
Dechter, R.: Bucket elimination: a unifying framework for reasoning. Artif. Intell. 113(1–2), 41–85 (1999)
Dechter, R.: Constraint Processing. Elsevier/Morgan Kaufmann (2003)
Dechter, R.: Reasoning with Probabilistic and Deterministic Graphical Models: Exact Algorithms. Synthesis Lectures on Artificial Intelligence and Machine Learning, 2nd edn. Morgan & Claypool Publishers (2019)
Dechter, R., Mateescu, R.: AND/OR search spaces for graphical models. Artif. Intell. 171(2–3), 73–106 (2007)
Erdeniz, S.P., Felfernig, A.: Cluster and learn: cluster-specific heuristics for graph coloring. In: PATAT 2018, pp. 401–404 (2018)
Freuder, E.C., Quinn, M.J.: Taking advantage of stable sets of variables in constraint satisfaction problems. In: IJCAI 1985, Los Angeles, CA, USA, August 1985, pp. 1076–1078 (1985)
Galassi, A., Lombardi, M., Mello, P., Milano, M.: Model agnostic solution of CSPs via deep learning: a preliminary study. In: van Hoeve, W.-J. (ed.) CPAIOR 2018. LNCS, vol. 10848, pp. 254–262. Springer, Cham (2018). https://doi.org/10.1007/978-3-319-93031-2_18
Gent, I.P., et al.: Learning when to use lazy learning in constraint solving. In: ECAI 2010, Lisbon, Portugal, 16–20 August 2010, Proceedings, vol. 215, pp. 873–878. IOS Press (2010)
George, A.: Nested dissection of a regular finite element mesh. SIAM J. Numer. Anal. 10(2), 345–363 (1973)
Gottlob, G., Greco, G., Leone, N., Scarcello, F.: Hypertree decompositions: questions and answers. In: Proceedings of the 35th ACM SIGMOD-SIGACT-SIGAI, PODS 2016, San Francisco, CA, USA, 26 June–01 July 2016, pp. 57–74. ACM (2016)
Guerri, A., Milano, M.: Learning techniques for automatic algorithm portfolio selection. In: ECAI 2004, Valencia, Spain, 22–27 August 2004, pp. 475–479. IOS Press (2004)
Huang, L., et al.: Branch and bound in mixed integer linear programming problems: a survey of techniques and trends. CoRR abs/2111.06257 (2021). https://arxiv.org/abs/2111.06257
Hurley, B., et al.: Multi-language evaluation of exact solvers in graphical model discrete optimization. Constraints Int. J. 21(3), 413–434 (2016). https://doi.org/10.1007/s10601-016-9245-y
Jégou, P., Terrioux, C.: Hybrid backtracking bounded by tree-decomposition of constraint networks. Artif. Intell. 146(1), 43–75 (2003)
Jégou, P., Terrioux, C.: Decomposition and good recording for solving Max-CSPs. In: ECAI 2004, Spain, 22–27 August 2004, pp. 196–200 (2004)
Jégou, P., Terrioux, C.: Combining restarts, nogoods and decompositions for solving CSPs. In: ECAI 2014, Czech Republic, 18–22 August 2014, vol. 263, pp. 465–470 (2014)
Jégou, P., Terrioux, C.: Combining restarts, nogoods and bag-connected decompositions for solving CSPs. Constraints Int. J. 22(2), 191–229 (2017). https://doi.org/10.1007/s10601-016-9248-8
Bayardo Jr., R.J., Miranker, D.P.: On the space-time trade-off in solving constraint satisfaction problems. In: IJCAI 1995, Québec, Canada, 20–25 August 1995, vol. 2, pp. 558–562 (1995)
Kitching, M., Bacchus, F.: Exploiting decomposition in constraint optimization problems. In: Stuckey, P.J. (ed.) CP 2008. LNCS, vol. 5202, pp. 478–492. Springer, Heidelberg (2008). https://doi.org/10.1007/978-3-540-85958-1_32
Larrosa, J., Meseguer, P., Sánchez-Fibla, M.: Pseudo-tree search with soft constraints. In: ECAI 2002, Lyon, France, July 2002, pp. 131–135. IOS Press (2002)
Larrosa, J., Schiex, T.: Solving weighted CSP by maintaining arc consistency. Artif. Intell. 159(1–2), 1–26 (2004)
Marinescu, R., Dechter, R.: AND/OR branch-and-bound for graphical models. In: IJCAI-2005, Edinburgh, Scotland, UK, 30 July–5 August 2005, pp. 224–229 (2005)
Otten, L., Dechter, R.: Anytime AND/OR depth-first search for combinatorial optimization. AI Commun. 25(3), 211–227 (2012)
Popescu, A., et al.: An overview of machine learning techniques in constraint solving. J. Intell. Inf. Syst. 58(1), 91–118 (2022). https://doi.org/10.1007/s10844-021-00666-5
Robertson, N., Seymour, P.D.: Graph minors. III. Planar tree-width. J. Comb. Theory Ser. B 36(1), 49–64 (1984)
Sánchez-Fibla, M., Allouche, D., de Givry, S., Schiex, T.: Russian doll search with tree decomposition. In: IJCAI 2009, California, USA, 11–17 July 2009, pp. 603–608 (2009)
Terrioux, C., Jégou, P.: Bounded backtracking for the valued constraint satisfaction problems. In: Rossi, F. (ed.) CP 2003. LNCS, vol. 2833, pp. 709–723. Springer, Heidelberg (2003). https://doi.org/10.1007/978-3-540-45193-8_48
Verfaillie, G., Lemaître, M., Schiex, T.: Russian doll search for solving constraint optimization problems. In: IAAI 1996, Portland, Oregon, USA, 4–8 August 1996, pp. 181–187 (1996)
Xu, H., Koenig, S., Kumar, T.K.S.: Towards effective deep learning for constraint satisfaction problems. In: Hooker, J. (ed.) CP 2018. LNCS, vol. 11008, pp. 588–597. Springer, Cham (2018). https://doi.org/10.1007/978-3-319-98334-9_38
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Petrova, A., Larrosa, J. (2023). On Learning When to Decompose Graphical Models. In: Sellmann, M., Tierney, K. (eds) Learning and Intelligent Optimization. LION 2023. Lecture Notes in Computer Science, vol 14286. Springer, Cham. https://doi.org/10.1007/978-3-031-44505-7_19
Download citation
DOI: https://doi.org/10.1007/978-3-031-44505-7_19
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-44504-0
Online ISBN: 978-3-031-44505-7
eBook Packages: Computer ScienceComputer Science (R0)