Evaluating Probabilistic Inference Techniques: A Question of “When,” not “Which”
Historically, it has been claimed that one inference algorithm or technique, say A, is better than another, say B, based on the running times on a test set of Bayesian networks. Recent studies have instead focusing on identifying situations where A is better than B, and vice versa. We review two cases where competing inference algorithms (techniques) have been successfully applied together in unison to exploit the best of both worlds. Next, we look at recent advances in identifying structure and semantics. Finally, we present possible directions of future work in exploiting structure and semantics for faster probabilistic inference.
KeywordsBayesian Network Directed Acyclic Graph Inference Algorithm Cost Measure Probabilistic Inference
Unable to display preview. Download preview PDF.
- 1.Butz, C.J., Chen, J., Konkel, K., Lingras, P.: A formal comparison of variable elimination and arc reversal in Bayesian network inference. Intell. Dec. Analysis 3(3), 173–180 (2009)Google Scholar
- 2.Butz, C.J., Konkel, K., Lingras, P.: Join tree propagation utilizing both arc reversal and variable elimination. Intl. J. Approx. Rea. (2010) (in press)Google Scholar
- 3.Butz, C.J., Hua, S.: An improved Lazy-AR approach to Bayesian network inference. In: Proc. of Canadian Conference on Artificial Intelligence, pp. 183–194 (2006)Google Scholar
- 5.Butz, C.J., Madsen, A.L., Williams, K.: Using four cost measures to determine arc reversal orderings. In: Proc. 11th European Conference on Symbolic and Quantitative Approaches to Reasoning with Uncertainty, pp. 110–121 (2011)Google Scholar
- 6.Butz, C.J., Yan, W.: The semantics of intermediate CPTs in variable elimination. In: Proc. 5th European Workshop on Probabilistic Graphical Models, pp. 41–49 (2010)Google Scholar
- 10.Hájek, P., Havránek, T., Jiroušek, R.: Uncertain Information Processing in Expert Systems. CRC Press, Ann Arbor (1992)Google Scholar
- 11.Hansen, P.F., Pedersen, P.T.: Risk analysis of conventional and solo watch keeping. Research Report, Department of Naval Architecture and Offshore Engineering, Technical University of Denmark (1998)Google Scholar
- 16.Madsen, A.L.: An empirical evaluation of possible variations of lazy propagation. In: Proc. 20th Conference on Uncertainty in Artificial Intelligence, pp. 366–373 (2004)Google Scholar
- 19.Olmsted, S.: On representing and solving decision problems, Ph.D. thesis, Department of Engineering Economic Systems. Stanford University, Stanford, CA (1983)Google Scholar
- 22.Shafer, G.: Probabilistic Expert Systems. Society for Industrial and Applied Mathematics (1996)Google Scholar
- 25.Zhang, N.L., Poole, D.: A simple approach to Bayesian network computations. In: Proc. 10th Canadian Conference on Artificial Intelligence, pp. 171–178 (1994)Google Scholar