Bayesian network learning with cutting planes

Abstract

The problem of learning the structure of Bayesian networks from complete discrete data with a limit on parent set size is considered. Learning is cast explicitly as an optimisation problem where the goal is to find a BN structure which maximises log marginal likelihood (BDe score). Integer programming, specifically the SCIP framework, is used to solve this optimisation problem. Acyclicity constraints are added to the integer program (IP) during solving in the form of cutting planes. Finding good cutting planes is the key to the success of the approach -the search for such cutting planes is effected using a sub-IP. Results show that this is a particularly fast method for exact BN learning.

References in corpus (3)

• A simple approach for finding the globally optimal Bayesian network structure
• On Sensitivity of the MAP Bayesian Network Structure to the Equivalent Sample Size Parameter
• Bayesian network learning by compiling to weighted MAX-SAT

Cited by in corpus (33)

• Optimization Problems for Machine Learning: A Survey
• Advances in Bayesian Network Learning using Integer Programming
• Learning Sparse Causal Models is not NP-hard
• Searching Multiregression Dynamic Models of Resting-State fMRI Networks Using Integer Programming
• Learning Structures of Bayesian Networks for Variable Groups
• An Improved Admissible Heuristic for Learning Optimal Bayesian Networks
• SparsityBoost: A New Scoring Function for Learning Bayesian Network Structure
• Advances in Learning Bayesian Networks of Bounded Treewidth
• Learning Chordal Markov Networks by Constraint Satisfaction
• Gradient-Based Neural DAG Learning
• Estimating causal structure using conditional DAG models
• An Empirical-Bayes Score for Discrete Bayesian Networks
• Evaluating Anytime Algorithms for Learning Optimal Bayesian Networks
• Evaluating structure learning algorithms with a balanced scoring function
• Improving Bayesian Network Structure Learning in the Presence of Measurement Error
• Consistent Second-Order Conic Integer Programming for Learning Bayesian Networks
• On the Convergence of Continuous Constrained Optimization for Structure Learning
• Integer Programming for Causal Structure Learning in the Presence of Latent Variables
• On Pruning for Score-Based Bayesian Network Structure Learning
• Switching Regression Models and Causal Inference in the Presence of Discrete Latent Variables
• Structure Learning in Bayesian Networks of Moderate Size by Efficient Sampling
• Exact Maximum Margin Structure Learning of Bayesian Networks
• Active Structure Learning of Bayesian Networks in an Observational Setting
• Towards Scalable Bayesian Learning of Causal DAGs
• Beyond Uniform Priors in Bayesian Network Structure Learning
• Structure learning in polynomial time: Greedy algorithms, Bregman information, and exponential families
• Efficient Learning of Bounded-Treewidth Bayesian Networks from Complete and Incomplete Data Sets
• Effective and efficient structure learning with pruning and model averaging strategies
• Greedy structure learning from data that contain systematic missing values
• ML4C: Seeing Causality Through Latent Vicinity
• Prequential MDL for Causal Structure Learning with Neural Networks
• Learning Bounded Treewidth Bayesian Networks with Thousands of Variables
• Incorporating Type II Error Probabilities from Independence Tests into Score-Based Learning of Bayesian Network Structure