Structure Learning in Bayesian Networks of Moderate Size by Efficient Sampling
arXiv:1501.04370
Abstract
We study the Bayesian model averaging approach to learning Bayesian network structures (DAGs) from data. We develop new algorithms including the first algorithm that is able to efficiently sample DAGs according to the exact structure posterior. The DAG samples can then be used to construct estimators for the posterior of any feature. We theoretically prove good properties of our estimators and empirically show that our estimators considerably outperform the estimators from the previous state-of-the-art methods.
51 pages
References in corpus (5)
- A simple approach for finding the globally optimal Bayesian network structure
- Data Analysis with Bayesian Networks: A Bootstrap Approach
- Bayesian network learning with cutting planes
- Bayesian structure learning using dynamic programming and MCMC
- An Improved Admissible Heuristic for Learning Optimal Bayesian Networks