Abstract
Deterministic search algorithm such as greedy search is apt to get into local maxima, and learning Bayesian networks (BNs) by stochastic search strategy attracts the attention of many researchers. In this paper we propose a BN learning approach, E-MDL, based on stochastic search, which evolves BN structures with an evolutionary algorithm and can not only avoid getting into local maxima, but learn BNs with hidden variables. When there exists incomplete data, E-MDL estimates the probability distributions over the local structures in BNs from incomplete data, then evaluates BN structures by a variant of MDL score. The experimental results on Alarm, Asia and an examplar network verify the validation of E-MDL algorithm.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
Heckerman, D.: Bayesian networks for data mining. Data Mining and Knowledge Discovery 1(1), 79–119 (1997)
Lam, W., Bacchus, F.: Learning Bayesian belief networks: An approach based on the MDL principle. Computational Intelligence 10, 269–293 (1994)
Zhang, S., Wang, X.: Algorithm for Bayesian networks structure learning based on information entropy. Mini-Micro Computer Systems 26(6), 983–986 (2005)
Chickering, D.: Learning equivalence classes of Bayesian-network structures. Journal of Machine Learning Research 2, 445–498 (2002)
Tsamardinos, I., Aliferis, C., Statnikov, A.: Time and sample efficient discovery of Markov blankets and direct causal relations. In: 9th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 673–678. ACM Press, USA (2003)
Spirtes, P., Glymour, C., Scheines, R.: Causation, prediction and search, 2nd edn. MIT Press, Cambridge, MA (2001)
Cheng, J., Greiner, R., Kelly, J., Bell, D.A., Liu, W.: Learning Bayesian networks from data: an information-theory based approach. The Artificial Intelligence Journal 137, 43–90 (2002)
Ioannis, T., Laura, E., Constantin, F.A.: The max-min hill-climbing Bayesian networks from data. Machine Learning 65(1), 31–78 (2006)
Larranaga, P.M.P., et al.: Structure learning of Bayesian networks by genetic algorithms: A performance analysis of control parameters. IEEE Journal on Pattern Analysis and Machine Intelligence 18(9), 912–926 (1996)
Myers, J., Laskey, K., Levitt, T.: Learning Bayesian networks from incomplete data with stochastic search algorithms. In: 15th Conf. on Uncertainty in Artificial Intelligence (1999)
Friedman, N.: The Bayesian structural EM algorithm. In: Fourteenth Conf. on Uncertainty in Artificial Intelligence (1998)
Tian, F., Lu, Y., Shi, C.: Learning Bayesian networks with hidden variables using the combination of EM and evolutionary algorithm. In: Cheung, D., Williams, G.J., Li, Q. (eds.) PAKDD 2001. LNCS (LNAI), vol. 2035, pp. 568–574. Springer, Heidelberg (2001)
Tian, F., Zhang, H., Lu, Y.: Learning Bayesian networks from incomplete data based on EMI method. In: Proceedings of ICDM 2003 Melbourne, Florida, USA, pp. 323–330 (2003)
Settimi, R., Smith, J.Q.: On the geometry of Bayesian graphical models with hidden variables. In: Proceedings of UAI 1998, Morgan Kaufmann, Madison, WI (1998)
Martin, J., VanLehn, K.: Discrete factor analysis: Learning hidden variables in Bayesian networks. Technical report, Department of Computer Science, University of Pittsburgh, PA (1995)
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 2007 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Tian, F., Zhang, Y., Wang, Z., Huang, H. (2007). Learning Bayesian Networks Using Evolutionary Algorithm and a Variant of MDL Score. In: Apolloni, B., Howlett, R.J., Jain, L. (eds) Knowledge-Based Intelligent Information and Engineering Systems. KES 2007. Lecture Notes in Computer Science(), vol 4694. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-74829-8_21
Download citation
DOI: https://doi.org/10.1007/978-3-540-74829-8_21
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-74828-1
Online ISBN: 978-3-540-74829-8
eBook Packages: Computer ScienceComputer Science (R0)