Copula Gaussian graphical models and their application to modeling functional disability data

TL;DR

This paper introduces copula Gaussian graphical models (CGGMs), a Bayesian framework for flexible graphical model selection that handles mixed data types, demonstrated on a functional disability dataset.

Contribution

The paper presents a novel semiparametric Bayesian approach for graphical models that accommodates binary, ordinal, and continuous variables simultaneously.

Findings

Successfully modeled a 16-dimensional functional disability dataset

Demonstrated broad applicability in social science and economics studies

Embedded graphical model selection within a semiparametric Gaussian copula

Abstract

We propose a comprehensive Bayesian approach for graphical model determination in observational studies that can accommodate binary, ordinal or continuous variables simultaneously. Our new models are called copula Gaussian graphical models (CGGMs) and embed graphical model selection inside a semiparametric Gaussian copula. The domain of applicability of our methods is very broad and encompasses many studies from social science and economics. We illustrate the use of the copula Gaussian graphical models in the analysis of a 16-dimensional functional disability contingency table.