Abstract
Many real-world domains exhibit rich relational structure and stochasticity and motivate the development of models that combine predicate logic with probabilities. These models describe probabilistic influences between attributes of objects that are related to each other through known domain relationships. To keep these models succinct, each such influence is considered independent of others, which is called the assumption of “independence of causal influences” (ICI). In this paper, we describe a language that consists of quantified conditional influence statements and captures most relational probabilistic models based on directed graphs. The influences due to different statements are combined using a set of combining rules such as Noisy-OR. We motivate and introduce multi-level combining rules, where the lower level rules combine the influences due to different ground instances of the same statement, and the upper level rules combine the influences due to different statements. We present algorithms and empirical results for parameter learning in the presence of such combining rules. Specifically, we derive and implement algorithms based on gradient descent and expectation maximization for different combining rules and evaluate them on synthetic data and on a real-world task. The results demonstrate that the algorithms are able to learn both the conditional probability distributions of the influence statements and the parameters of the combining rules.
Similar content being viewed by others
References
Altendorf, E.E., Restificar, A.C., Dietterich, T.G.: Learning from sparse data by exploiting monotonicity constraints. In: Proceedings of UAI 05 (2005)
Binder, J., Koller, D., Russell, S., Kanazawa, K.: Adaptive probabilistic networks with hidden variables. Mach. Learn. 29(2–3), 213–244 (1997) ISSN 0885-6125
Dempster, A.P., Laird, N.M., Rubin, D.B.: Maximum likelihood from incomplete data via the EM algorithm. J. R. Stat. Soc. B 39, 185–197 (1977)
Díez, F.J., Galán, S.F.: Efficient computation for the noisy MAX. Int. J. Approx. Reason. 18, 165–177 (2003)
Domingos, P., Richardson, M.: Markov logic: a unifying framework for statistical relational learning. In: Proceedings of the SRL Workshop in ICML, Banff, July 2004
Dragunov, A.N., Dietterich, T.G., Johnsrude, K., McLaughlin, M., Li, L., Herlocker, J.L.: Tasktracer: a desktop environment to support multi-tasking knowledge workers. In: Proceedings of IUI, San Diego, January 2005
Fierens, D., Blockeel, H., Bruynooghe, M., Ramon, J.: Logical Bayesian networks and their relation to other probabilistic logical models. In: Proceedings of ILP, Bonn, 10–13 August 2005
Getoor, L., Grant, J.: PRL: a probabilistic relational language. Mach. Learn. 62(1–2), 7–31 (2006)
Getoor, L., Taskar, B.: Introduction to Statistical Relational Learning. MIT, Cambridge (2007)
Getoor, L., Friedman, N., Koller, D., Pfeffer, A.: Learning probabilistic relational models. In: Dzeroski, S., Lavrac, N. (eds.) Relational Data Mining. Springer, New York (2001)
Hastie, T., Tibshirani, R., Friedman, J.: The Elements of Statistical Learning. Springer, New York (2001)
Heckerman, D., Breese, J.S.: Causal independence for probability assessment and inference using Bayesian networks. Technical Report MSR-TR-94-08, Microsoft Research (1994)
Heckerman, D., Meek, C., Koller, D.: Probabilistic models for relational data. Technical Report MSR-TR-2004-30, March (2004)
Jaeger, M.: Relational Bayesian networks. In: Proceedings of UAI-97, Providence, 1–3 August 1997
Jaeger, M.: Parameter learning for relational Bayesian networks. In: Proceedings of the International Conference in Machine Learning, Corvalis, 20–24 June 2007
Kersting, K., De Raedt, L.: Bayesian logic programs. In: Proceedings of the Work-in-Progress Track at the 10th International Conference on Inductive Logic Programming, London, 24–27 July 2000
Kersting, K., De Raedt, L.: Adaptive Bayesian logic programs. In: Proceedings of the ILP ’01, pp. 104–117. Springer, New York (2001)
Koller, D., Pfeffer, A.: Learning probabilities for noisy first-order rules. In: IJCAI, pp. 1316–1323. Nagoya, 23–29 August 1997
Lafferty, J., McCallum, A., Pereira, F.: Conditional random fields: probabilistic models for segmenting and labeling sequence data. In: Proc. 18th International Conf. on Machine Learning, pp. 282–289 (2001)
Laskey, K.B.: MEBN: a language for first-order Bayesian knowledge bases. Artif. Intell. 172(2–3), 140–178 (2008)
Muggleton, S.: Stochastic logic programs. In: Advances in Inductive Logic Programming, pp. 254–264 (1996)
Natarajan, S., Tadepalli, P., Altendorf, E., Dietterich, T.G., Fern, A., Restificar, A.: Learning first-order probabilistic models with combining rules. In: Proceedings of the International Conference in Machine Learning, Bonn, 7–11 August 2005
Natarajan, S., Tadepalli, P., Fern, A.: A relational hierarchical model for decision-theoretic assistance. In: Proceedings of 17th Annual International Conference on Inductive Logic Programming, Corvallis, 19–21 June 2007
Neville, J., Jensen, D., Friedland, L., Hay, M.: Learning relational probability trees. In: KDD ’03: proceedings of the ninth ACM SIGKDD international conference on knowledge discovery and data mining, pp. 625–630. Washington, DC, 24–27 August 2003
Ngo, L., Haddawy, P.: Probabilistic logic programming and Bayesian networks. In: Proceedings ACSC95, Pathumthani, 11–13 December 1995
Poole, D.: Probabilistic Horn abduction and Bayesian networks. Artif. Intell. 64(1), 81–129 (1993)
Sato, T., Kameya, Y.: Parameter learning of logic programs for symbolic-statistical modeling. J. Artif. Intell. Res. 15, 391–454 (2001)
Vomlel, J.: Noisy-or classifier: research articles. Int. J. Intell. Syst. 21(3), 381–398 (2006)
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Natarajan, S., Tadepalli, P., Dietterich, T.G. et al. Learning first-order probabilistic models with combining rules. Ann Math Artif Intell 54, 223–256 (2008). https://doi.org/10.1007/s10472-009-9138-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10472-009-9138-5