Peruvemba Ramaswamy, V., & Szeider, S. (2022). Learning Fast-Inference Bayesian Networks. In Advances in Neural Information Processing Systems 34 (NeurIPS 2021). 35th conference on neural information processing systems (NeurIPS 2021), Unknown. https://doi.org/10.34726/4023
We propose new methods for learning Bayesian networks (BNs) that reliably support fast inference. We utilize maximum state space size as a more fine-grained measure for the BN's reasoning complexity than the standard treewidth measure, thereby accommodating the possibility that variables range over domains of different sizes. Our methods combine heuristic BN structure learning algorithms with the recently introduced MaxSAT-powered local improvement method (Peruvemba Ramaswamy and Szeider, AAAI'21). Our experiments show that our new learning methods produce BNs that support significantly faster exact probabilistic inference than BNs learned with treewidth bounds.
en
Project title:
SAT-Basierende lokale Verbesserungsmethoden: P32441-N35 (FWF Fonds zur Förderung der wissenschaftlichen Forschung (FWF)) Revealing and Utilizing the Hidden Structure for Solving Hard Problems in AI: ICT19-065 (WWTF Wiener Wissenschafts-, Forschu und Technologiefonds)
