Learning and interpreting asymmetry-labeled DAGs: a case study on COVID-19 fear

TL;DR

This paper introduces efficient structural learning algorithms for asymmetry-labeled DAGs, extending Bayesian networks to better model asymmetric dependencies, with demonstrated effectiveness on COVID-19 fear data.

Contribution

The paper presents novel, efficient algorithms for learning asymmetry-labeled DAGs, enabling more interpretable modeling of asymmetric dependencies in complex data.

Findings

Algorithms are computationally efficient.

Effective in real-world COVID-19 fear data analysis.

Enhanced interpretability of dependence structures.

Abstract

Bayesian networks are widely used to learn and reason about the dependence structure of discrete variables. However, they are only capable of formally encoding symmetric conditional independence, which in practice is often too strict to hold. Asymmetry-labeled DAGs have been recently proposed to both extend the class of Bayesian networks by relaxing the symmetric assumption of independence and denote the type of dependence existing between the variables of interest. Here, we introduce novel structural learning algorithms for this class of models which, whilst being efficient, allow for a straightforward interpretation of the underlying dependence structure. A comprehensive computational study highlights the efficiency of the algorithms. A real-world data application using data from the Fear of COVID-19 Scale collected in Italy showcases their use in practice.