Learning Bayesian Networks: A Unification for Discrete and Gaussian Domains

Abstract

We examine Bayesian methods for learning Bayesian networks from a combination of prior knowledge and statistical data. In particular, we unify the approaches we presented at last year's conference for discrete and Gaussian domains. We derive a general Bayesian scoring metric, appropriate for both domains. We then use this metric in combination with well-known statistical facts about the Dirichlet and normal--Wishart distributions to derive our metrics for discrete and Gaussian domains.

This version has improved pointers to the literature

References in corpus (1)

• A Transformational Characterization of Equivalent Bayesian Network Structures

Cited by in corpus (14)

• Being Bayesian about Network Structure
• Parameter Priors for Directed Acyclic Graphical Models and the Characterization of Several Probability Distributions
• Prediction, Expectation, and Surprise: Methods, Designs, and Study of a Deployed Traffic Forecasting Service
• Gaussian Process Networks
• Addendum on the scoring of Gaussian directed acyclic graphical models
• Parameter Priors for Directed Acyclic Graphical Models and the Characterization of Several Probability Distributions
• A Case Study in Knowledge Discovery and Elicitation in an Intelligent Tutoring Application
• Likelihoods and Parameter Priors for Bayesian Networks
• Learning Hidden Markov Models with Geometrical Constraints
• Hybrid Probabilistic Inference with Logical Constraints: Tractability and Message Passing
• Scaling up Hybrid Probabilistic Inference with Logical and Arithmetic Constraints via Message Passing
• Learning Bayesian Networks from Ordinal Data
• Deep Neural Programs for Adaptive Control in Cyber-Physical Systems
• Bayesian Model Averaging for Data Driven Decision Making when Causality is Partially Known