MaxSAT-Based Cutting Planes for Learning Graphical Models

Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9075)

Abstract

A way of implementing domain-specific cutting planes in branch-and-cut based Mixed-Integer Programming (MIP) solvers is through solving so-called sub-IPs, solutions of which correspond to the actual cuts. We consider the suitability of using Maximum satisfiability solvers instead of MIP for solving sub-IPs. As a case study, we focus on the problem of learning optimal graphical models, namely, Bayesian and chordal Markov network structures.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Achterberg, T.: Constrained Integer Programming. Ph.D. Thesis, TU Berlin (2007)Google Scholar
  2. 2.
    Ansotegui, C., Bonet, M.L., Levy, J.: SAT-based MaxSAT algorithms. Artificial Intelligence 196, 77–105 (2013)CrossRefMATHMathSciNetGoogle Scholar
  3. 3.
    Bartlett, M., Cussens, J.: Advances in Bayesian network learning using integer programming. In: Proc. UAI, pp. 182–191. AUAI Press (2013)Google Scholar
  4. 4.
    de Campos, C.P., Zeng, Z., Ji, Q.: Structure learning of Bayesian networks using constraints. In: Proc. ICML, pp. 113–120. ACM (2009)Google Scholar
  5. 5.
    Corander, J., Janhunen, T., Rintanen, J., Nyman, H., Pensar, J.: Learning chordal Markov networks by constraint satisfaction. In: Proc. NIPS, pp. 1349–1357. Curran Associates, Inc. (2013)Google Scholar
  6. 6.
    Cussens, J.: Bayesian network learning with cutting planes. In: Proc. UAI, pp. 153–160. AUAI Press (2011)Google Scholar
  7. 7.
    Cussens, J., Bartlett, M.: GOBNILP 1.4.1 user/developer manual (2013)Google Scholar
  8. 8.
    Davies, J., Bacchus, F.: Exploiting the power of mip solvers in maxsat. In: Järvisalo, M., Van Gelder, A. (eds.) SAT 2013. LNCS, vol. 7962, pp. 166–181. Springer, Heidelberg (2013) CrossRefGoogle Scholar
  9. 9.
    Heckerman, D., Geiger, D., Chickering, D.M.: Learning Bayesian networks: the combination of knowledge and statistical data. Machine Learning 20, 197–243 (1995)MATHGoogle Scholar
  10. 10.
    Jaakkola, T., Sontag, D., Globerson, A., Meila, M.: Learning Bayesian network structure using LP relaxations. In: Proc. AISTATS, pp. 358–365. JMLR.org (2010)Google Scholar
  11. 11.
    Kangas, K., Niinimäki, T., Koivisto, M.: Learning chordal Markov networks by dynamic programming. In: Proc. NIPS, pp. 2357–2365. Curran Associates, Inc. (2014)Google Scholar
  12. 12.
    Koivisto, M., Sood, K.: Exact Bayesian structure discovery in Bayesian networks. Journal of Machine Learning Research, 549–573 (2004)Google Scholar
  13. 13.
    Li, C.M., Manyà, F.: MaxSAT, hard and soft constraints. In: Handbook of Satisfiability, Frontiers in Artificial Intelligence and Applications, vol. 185, chap. 19, pp. 613–631. IOS Press (2009)Google Scholar
  14. 14.
    Malone, B., Kangas, K., Järvisalo, M., Koivisto, M., Myllymäki, P.: Predicting the hardness of learning Bayesian networks. In: Proc. AAAI, pp. 1694–1700. AAAI Press (2014)Google Scholar
  15. 15.
    Morgado, A., Heras, F., Liffiton, M.H., Planes, J., Marques-Silva, J.: Iterative and core-guided MaxSAT solving: a survey and assessment. Constraints 18(4), 478–534 (2013)CrossRefMathSciNetGoogle Scholar
  16. 16.
    Ott, S., Miyano, S.: Finding optimal gene networks using biological constraints. Genome Informatics 14, 124–133 (2003)Google Scholar
  17. 17.
    Pearl, J.: Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference. Morgan Kaufmann Publishers Inc. (1988)Google Scholar
  18. 18.
    Silander, T., Myllymäki, P.: A simple approach for finding the globally optimal Bayesian network structure. In: Proc. UAI, pp. 445–452. AUAI Press (2006)Google Scholar
  19. 19.
    Yuan, C., Malone, B.: Learning optimal Bayesian networks: a shortest path perspective. Journal of Artificial Intelligence Research 48, 23–65 (2013)MATHMathSciNetGoogle Scholar

Copyright information

© Springer International Publishing Switzerland 2015

Authors and Affiliations

  1. 1.Helsinki Institute for Information Technology, Department of Computer ScienceUniversity of HelsinkiHelsinkiFinland

Personalised recommendations