B1638
Title: Bayesian structure learning in undirected graphical models: Review and empirical comparisons
Authors: Lucas Vogels - University of Amsterdam (Netherlands) [presenting]
Reza Mohammadi - University of Amsterdam (Netherlands)
Ilker Birbil - University of Amsterdam (Netherlands)
Marit Schoonhoven - University of Amsterdam (Netherlands)
Abstract: Graphical models are an elegant way to depict the conditional dependencies among variables using a graph. Bayesian structure learning is the area occupied with revealing the structure of this graph using Bayesian methods. Although multiple solution methods have been proposed in this field over the last decade, no comprehensive review or empirical comparison is available. This review is presented. We will list and classify all methods, compare their performance in a simulation study, and give suggestions for future research.
