摘要: Most Relevant Explanation (MRE) is an inference task in Bayesian networks that finds the most relevant partial instantiation of target variables as explanation for given evidence by maximizing Generalized Bayes Factor (GBF). No exact MRE algorithm has been developed previously except exhaustive search. This paper fills void introducing two Breadth-First Branch-and-Bound (BFBnB) algorithms solving based on novel upper bounds GBF. One bound created decomposing computation GBF using a blanket decomposition variables. The other improves first ways. to split blankets are too large converting auxiliary nodes into pseudo-targets so scale problems. perform summations instead maximizations some each blanket. Our empirical evaluations show proposed BFBnB make tractable could not be solved previously.
sci-hub.st 参考文章(38)
Marek J. Druzdzel, Max Henrion, Qualtitative propagation and scenario-based scheme for exploiting probabilistic reasoning uncertainty in artificial intelligence. pp. 17- 32 ,(1990)
M. Julia Flores, José A. Gámez, Serafín Moral, Abductive Inference in Bayesian Networks: Finding a Partition of the Explanation Space Lecture Notes in Computer Science. pp. 63- 75 ,(2005) , 10.1007/11518655_7
Adnan Darwiche, James D. Park, Solving MAP exactly using systematic search uncertainty in artificial intelligence. pp. 459- 468 ,(2002)
Jayant Kalagnanam, Max Henrion, A comparison of decision alaysis and expert rules for sequential diagnosis uncertainty in artificial intelligence. ,vol. 9, pp. 271- 282 ,(1990) , 10.1016/B978-0-444-88650-7.50026-3
Johan Kwisthout, Most inforbable explanations: finding explanations in bayesian networks that are both probable and informative european conference on symbolic and quantitative approaches to reasoning and uncertainty. pp. 328- 339 ,(2013) , 10.1007/978-3-642-39091-3_28
Ingo A. Beinlich, H. J. Suermondt, R. Martin Chavez, Gregory F. Cooper, The ALARM Monitoring System: A Case Study with two Probabilistic Inference Techniques for Belief Networks artificial intelligence in medicine in europe. pp. 247- 256 ,(1989) , 10.1007/978-3-642-93437-7_28
Nir Friedman, Daniel L. Koller, Probabilistic graphical models : principles and techniques The MIT Press. ,(2009)
Changhe Yuan, Eric A. Hansen, Efficient computation of jointree bounds for systematic MAP search international joint conference on artificial intelligence. pp. 1982- 1989 ,(2009)
Adnan Darwiche, Modeling and Reasoning with Bayesian Networks ,(2014)
M.L. Wessels, L. van der Gaag, Selective evidence gathering for diagnostic belief networks ,(1993)