Abstract
Bayesian network structure learning (BNSL) is the problem of finding a BN structure which best explains a dataset. Score-based learning assigns a score to each network structure. The goal is to find the structure which optimizes the score. We review two recent studies of empirical behavior of BNSL algorithms.
The score typically reflects fit to a training dataset; however, models which fit training data well may generalize poorly. Thus, it is not clear that finding an optimal network is worthwhile. We review a comparison of exact and approximate search techniques. Sometimes, approximate algorithms suffice; for complex datasets, the optimal algorithms produce better networks.
BNSL is known to be NP-hard, so exact solvers prune the search space using heuristics. We next review problem-dependent characteristics which affect their efficacy. Empirical results show that machine learning techniques based on these characteristics can often be used to accurately predict the algorithms’ running times.
B. Malone—This paper is based on Malone et al. (2014, 2015), with co-authors Matti Järvisalo, Petri Myllymäki, Kusta Kangas and Mikko Koiviso from HIIT and the Department of Computer Science at the University of Helsinki.
Notes
- 1.
This work assumes s is score-equivalent (Heckerman et al. 1995).
- 2.
The datasets are available at http://bnportfolio.cs.helsinki.fi/.
- 3.
Our results were not very sensitive to the scoring function, except its effect on the number of CPSs, so our results generalize to other decomposable scores.
References
Bartlett, M., Cussens, J.: Integer linear programming for the Bayesian network structure learning problem. In: Artificial Intelligence (2015)
Chickering, D.M.: Learning Bayesian networks is NP-complete. In: Fisher, D., Lenz, H.-J. (eds.) Learning from Data: Artificial Intelligence and Statistics V. Lecture Notes in Statistics, vol. 112, pp. 121–130. Springer, New York (1996)
Chickering, D.M.: Learning equivalence classes of Bayesian-network structures. J. Mach. Learn. Res. 2, 445–498 (2002)
Chow, C., Liu, C.: Approximating discrete probability distributions with dependence trees. IEEE Trans. Inf. Theory 14(3), 462–467 (1968)
Cooper, G.F., Herskovits, E.: A Bayesian method for the induction of probabilistic networks from data. Mach. Learn. 9, 309–347 (1992)
Cussens, J., Bartlett, M., Jones, E.M., Sheehan, N.A.: Maximum likelihood pedigree reconstruction using integer linear programming. Genet. Epidemiol. 37(1), 69–83 (2013)
de Campos, C.P., Ji, Q.: Efficient learning of Bayesian networks using constraints. J. Mach. Learn. Res. 12, 663–689 (2011)
de Campos, L.M., Huete, J.F.: A new approach for learning belief networks using independence criteria. Int. J. Approximate Reasoning 24(1), 11–37 (2000)
Fan, X., Yuan, C., Malone, B.: Tightening bounds for Bayesian network structure learning. In: Proceedings of the 28th AAAI Conference on Artificial Intelligence (2014)
Friedman, N., Nachman, I., Peer, D.: Learning Bayesian network structure from massive datasets: the “sparse candidate” algorithm. In: Proceedings 13th Conference on Uncertainty in Artificial Intelligence (1999)
Glover, F.: Tabu search: a tutorial. Interfaces 20(4), 74–94 (1990)
Heckerman, D., Geiger, D., Chickering, D.M.: Learning Bayesian networks: the combination of knowledge and statistical data. Mach. Learn. 20, 197–243 (1995)
Koivisto, M., Sood, K.: Exact Bayesian structure discovery in Bayesian networks. J. Mach. Learn. Res. 5, 549–573 (2004)
Leyton-Brown, K., Nudelman, E., Shoham, Y.: Learning the empirical hardness of optimization problems: the case of combinatorial auctions. In: Hentenryck, P. (ed.) CP 2002. LNCS, vol. 2470, pp. 556–572. Springer, Heidelberg (2002)
Malone, B., Järvisalo, M., Myllymäki, P.: Impact of learning strategies on the qualpacking Bayesian networks: an empirical evaluation. In: Proceedings of the 31st Conference on Uncertainty in Artificial Intelligence (2015)
Malone, B., Kangas, K., Järvisalo, M., Koivisto, M., Myllymäki, P.: Predicting the hardness of learning Bayesian networks. In: Proceedings of the 28th AAAI Conference on Artificial Intelligence (2014)
Malone, B., Yuan, C.: Evaluating anytime algorithms for learning optimal Bayesian networks. In: Proceedings of the 29th Conference on Uncertainty in Artificial Intelligence (2013)
Mitchell, T.: Machine Learning. McGraw-Hill, New York (1997)
Moore, A., Wong, W.-K.: Optimal reinsertion: a new search operator for accelerated and more accurate Bayesian network structure learning. In: Proceedings of the 20th International Conference on Machine Learning, pp. 552–559 (2003)
Ott, S., Imoto, S., Miyano, S.: Finding optimal models for small gene networks. In: Proceedings of the Pacific Symposium on Biocomputing (2004)
Parviainen, P., Koivisto, M.: Exact structure discovery in Bayesian networks with less space. In: Proceedings of the Twenty-Fifth Conference on Uncertainty in Artificial Intelligence, Montreal, Quebec, Canada. AUAI Press (2009)
Pearl, J.: Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference. Morgan Kaufmann Publishers Inc., San Mateo (1988)
Quinlan, J.R.: Simplifying decision trees. Int. J. Man Mach. Stud. 27, 221–234 (1987)
Rice, J.R.: The algorithm selection problem. Adv. Comput. 15, 65–118 (1976)
Russell, S.J., Norvig, P.: Artificial Intelligence: A Modern Approach. Pearson Education, Upper Saddle River (2003)
Silander, T., Myllymäki, P.: A simple approach for finding the globally optimal Bayesian network structure. In: Proceedings of the 22nd Conference on Uncertainty in Artificial Intelligence (2006)
Silander, T., Roos, T., Kontkanen, P., Myllymäki, P.: Factorized normalized maximum likelihood criterion for learning Bayesian network structures. In: Proceedings of the 4th European Workshop on Probabilistic Graphical Models (2008)
Spirtes, P., Glymour, C., Schemes, R.: Causation, Prediction, and Search, 2nd edn. MIT Press, Cambridge (2000)
Suzuki, J.: Learning Bayesian belief networks based on the MDL principle: an efficient algorithm using the branch and bound technique. IEICE Trans. Inf. Syst. E82–D(2), 356–367 (1999)
Teyssier, M., Koller, D.: Ordering-based search: a simple and effective algorithm for learning Bayesian networks. In: Proceedings of the 21st Conference on Uncertainty in Artificial Intelligence (2005)
Tian, J.: A branch-and-bound algorithm for MDL learning Bayesian networks. In: Proceedings of the 16th Conference on Uncertainty in Artificial Intelligence (2000)
Tsamardinos, I., Brown, L., Aliferis, C.: The max-min hill-climbing Bayesian network structure learning algorithm. Mach. Learn. 65, 31–78 (2006)
van Beek, P., Hoffmann, H.-F.: Machine learning of Bayesian networks using constraint programming. In: Pesant, G. (ed.) CP 2015. LNCS, vol. 9255, pp. 429–445. Springer, Heidelberg (2015)
Yuan, C., Malone, B.: Learning optimal Bayesian networks: a shortest path perspective. J. Artif. Intell. Res. 48, 23–65 (2013)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2015 Springer International Publishing Switzerland
About this paper
Cite this paper
Malone, B. (2015). Empirical Behavior of Bayesian Network Structure Learning Algorithms. In: Suzuki, J., Ueno, M. (eds) Advanced Methodologies for Bayesian Networks. AMBN 2015. Lecture Notes in Computer Science(), vol 9505. Springer, Cham. https://doi.org/10.1007/978-3-319-28379-1_8
Download citation
DOI: https://doi.org/10.1007/978-3-319-28379-1_8
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-28378-4
Online ISBN: 978-3-319-28379-1
eBook Packages: Computer ScienceComputer Science (R0)