A Multivariate Discretization Method for Learning Bayesian Networks from Mixed Data

TL;DR

This paper introduces a multivariate discretization method for learning Bayesian networks from mixed data, dynamically adjusting discretizations based on network structure and data interactions.

Contribution

It presents a novel Bayesian scoring-based discretization technique that considers variable interactions during BN structure learning from mixed data.

Findings

Improved discretization accuracy for mixed data

Enhanced Bayesian network learning performance

Dynamic adjustment of discretization during structure search

Abstract

In this paper we address the problem of discretization in the context of learning Bayesian networks (BNs) from data containing both continuous and discrete variables. We describe a new technique for <EM>multivariate</EM> discretization, whereby each continuous variable is discretized while taking into account its interaction with the other variables. The technique is based on the use of a Bayesian scoring metric that scores the discretization policy for a continuous variable given a BN structure and the observed data. Since the metric is relative to the BN structure currently being evaluated, the discretization of a variable needs to be dynamically adjusted as the BN structure changes.