Bayesian network model selection using integer programming
With complete data and appropriately chosen parameter priors the problem of finding a Bayesian network with maximal log marginal likelihood (LML) becomes a purely discrete problem: search for a directed acyclic graph (DAG) with maximal LML. We solve this problem of discrete optimisation using integer linear programming (ILP) with the SCIP (Solving Constraint Integer Programming) framework. In many cases this allows us to solve the problem: we find a DAG which we know to have maximal LML. Also using ILP allows prior knowledge, such as known conditional independence relations, to be expressed as constraints on DAG structure The key to efficient solving is to add certain linear constraints ruling out cyclic digraphs during the search. I will report on the successes and limitations of this approach and discuss future directions.
Date: 4 June 2015, 14:15 (Thursday, 6th week, Trinity 2015)
Venue: 1 South Parks Road, 1 South Parks Road OX1 3TG
Venue Details: Lecture Theatre, Department of Statistics
Speaker: Dr James Cussens (University of York)
Organising department: Department of Statistics
Organisers: Professor Yee Whye Teh (Department of Statistics, University of Oxford), Dr Robin Evans (Department of Statistics, University of Oxford)
Organiser contact email address: lane@stats.ox.ac.uk
Part of: Statistics, Applied Probability & Operational Research Seminars
Booking required?: Not required
Audience: Members of the University only
Editor: Beverley Lane