DeepAI AI Chat
Log In Sign Up

Robust learning Bayesian networks for prior belief

by   Maomi Ueno, et al.

Recent reports have described that learning Bayesian networks are highly sensitive to the chosen equivalent sample size (ESS) in the Bayesian Dirichlet equivalence uniform (BDeu). This sensitivity often engenders some unstable or undesirable results. This paper describes some asymptotic analyses of BDeu to explain the reasons for the sensitivity and its effects. Furthermore, this paper presents a proposal for a robust learning score for ESS by eliminating the sensitive factors from the approximation of log-BDeu.


Learning networks determined by the ratio of prior and data

Recent reports have described that the equivalent sample size (ESS) in a...

Dirichlet Bayesian Network Scores and the Maximum Entropy Principle

A classic approach for learning Bayesian networks from data is to select...

A Branch-and-Bound Algorithm for MDL Learning Bayesian Networks

This paper extends the work in [Suzuki, 1996] and presents an efficient ...

On Bayesian Dirichlet Scores for Staged Trees and Chain Event Graphs

Chain event graphs (CEGs) are a recent family of probabilistic graphical...

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

BDeu marginal likelihood score is a popular model selection criterion fo...

Bayesian Networks, Total Variation and Robustness

Now that Bayesian Networks (BNs) have become widely used, an appreciatio...

Parallel Belief Revision

This paper describes a formal system of belief revision developed by Wol...