Abstract:Doubly-intractable posterior distributions arise in many applications of statistics concerned with discrete and dependent data, including physics, spatial statistics, machine learning, the social sciences, and other fields. A specific example is psychometrics, which has adapted high-dimensional Ising models from machine learning, with a view to studying the interactions among binary item responses in educational assessments. To estimate high-dimensional Ising models from educational assessment data, $\ell_1$-penalized nodewise logistic regressions have been used. Theoretical results in high-dimensional statistics show that $\ell_1$-penalized nodewise logistic regressions can recover the true interaction structure with high probability, provided that certain assumptions are satisfied. Those assumptions are hard to verify in practice and may be violated, and quantifying the uncertainty about the estimated interaction structure and parameter estimators is challenging. We propose a Bayesian approach that helps quantify the uncertainty about the interaction structure and parameters without requiring strong assumptions, and can be applied to Ising models with thousands of parameters. We demonstrate the advantages of the proposed Bayesian approach compared with $\ell_1$-penalized nodewise logistic regressions by simulation studies and applications to small and large educational data sets with up to 2,485 parameters. Among other things, the simulation studies suggest that the Bayesian approach is more robust against model misspecification due to omitted covariates than $\ell_1$-penalized nodewise logistic regressions.

Bayesian finite mixtures of Ising models

Bayesian mixtures of common factor analyzers: Model, variational inference, and applications

Bayesian estimation and prediction for certain mixtures

Model-based clustering and classification using mixtures of multivariate skewed power exponential distributions

Bayesian inference for finite mixtures of univariate and multivariate skew-normal and skew-t distributions

Bayesian mixture modeling for multivariate conditional distributions

Bayesian Mixtures Models with Repulsive and Attractive Atoms

The Infinite Student'S T-Mixture For Robust Modeling

Bayesian Model Selection for High-Dimensional Ising Models, With Applications to Educational Data

Bayesian analysis for finite mixture in non-recursive non-linear structural equation models.

Mixture Models With a Prior on the Number of Components

Simultaneous Bayesian Clustering and Model Selection with Mixture of Robust Factor Analyzers

A Mean Field Games model for finite mixtures of Bernoulli and Categorical distributions

Local Bayesian Dirichlet mixing of imperfect models

A Bayesian approach for clustering skewed data using mixtures of multivariate normal-inverse Gaussian distributions

Bayesian Mixture Models With Focused Clustering for Mixed Ordinal and Nominal Data

Bayesian joint analysis of heterogeneous- and skewed-longitudinal data and a binary outcome, with application to AIDS clinical studies

Bayesian mixture models (in)consistency for the number of clusters

Overfitting Bayesian Mixtures of Factor Analyzers with an Unknown Number of Components

Nonparametric Bayesian Methods (Dirichlet Process Mixtures)

Nearest Neighbor Dirichlet Mixtures