Scale Up Bayesian Networks Learning Dissertation Proposal

Bayesian networks are widely used graphical models which represent uncertain relations between the random variables in a domain compactly and intuitively. The first step of applying Bayesian networks to real-word problems typically requires building the network structure. Among others, optimal structure learning via score-and-search has become an active research topic in recent years. In this context, a scoring function is used to measure the goodness of fit of a structure to the data, and the goal is to find the structure which optimizes the scoring function. The problem has been shown to be NP-hard. This proposal focuses on scaling up structure learning algorithms. Two potential directions are preliminarily explored. One is to tighten the bounds; we tighten the lower bound by using more informed variable groupings when creating the heuristics. and tighten the upper bound using anytime learning algorithm. The other is to prune the search space by extracting constraints directly from data. Our preliminary results show that the two directions can significantly improve the efficiency and scalability of heuristics search-based structure learning algorithms, which gives us strong reasons to follow these directions in the future. Professor: Dr. Changhe Yuan Signature: Date: 03-11-2015 Professor: Dr. Liang Huang Signature: Date: 03-11-2015 Professor: Dr. Andrew Rosenberg Signature: Date: 03-11-2015 Professor: Dr.John Paisley Signature: Date: 03-11-2015 Advisor: Dr. Changhe Yuan 1