On Sensitivity of the MAP Bayesian Network Structure to the Equivalent Sample Size Parameter

Abstract

BDeu marginal likelihood score is a popular model selection criterion for selecting a Bayesian network structure based on sample data. This non-informative scoring criterion assigns same score for network structures that encode same independence statements. However, before applying the BDeu score, one must determine a single parameter, the equivalent sample size alpha. Unfortunately no generally accepted rule for determining the alpha parameter has been suggested. This is disturbing, since in this paper we show through a series of concrete experiments that the solution of the network structure optimization problem is highly sensitive to the chosen alpha parameter value. Based on these results, we are able to give explanations for how and why this phenomenon happens, and discuss ideas for solving this problem.

Appears in Proceedings of the Twenty-Third Conference on Uncertainty in Artificial Intelligence (UAI2007)

References in corpus (1)

• A simple approach for finding the globally optimal Bayesian network structure

Cited by in corpus (10)

• Bayesian network learning with cutting planes
• Learning Bayesian Network Equivalence Classes with Ant Colony Optimization
• Labeled Directed Acyclic Graphs: a generalization of context-specific independence in directed graphical models
• Causal datasheet: An approximate guide to practically assess Bayesian networks in the real world
• An Empirical-Bayes Score for Discrete Bayesian Networks
• Learning the Bayesian Network Structure: Dirichlet Prior versus Data
• A Framework for Assurance of Medication Safety using Machine Learning
• A Bayesian Dynamic Graphical Model for Recurrent Events in Public Health
• Exact Learning Augmented Naive Bayes Classifier
• Beyond Uniform Priors in Bayesian Network Structure Learning