Bayesian Semiparametric Hidden Markov Tensor Partition Models for Longitudinal Data with Local Variable Selection

TL;DR

This paper introduces a Bayesian semiparametric hidden Markov tensor model for analyzing longitudinal data with high-dimensional categorical covariates, enabling flexible variable selection and time-varying effects.

Contribution

It develops a novel hidden Markov tensor decomposition technique that allows covariate effects to vary over time and across dependent partitions, with theoretical guarantees and practical algorithms.

Findings

Effective variable selection in high-dimensional settings

Flexible modeling of time-varying covariate effects

Strong empirical performance in synthetic and real data

Abstract

We present a flexible Bayesian semiparametric mixed model for longitudinal data analysis in the presence of potentially high-dimensional categorical covariates. Building on a novel hidden Markov tensor decomposition technique, our proposed method allows the fixed effects components to vary between dependent random partitions of the covariate space at different time points. The mechanism not only allows different sets of covariates to be included in the model at different time points but also allows the selected predictors' influences to vary flexibly over time. Smooth time-varying additive random effects are used to capture subject specific heterogeneity. We establish posterior convergence guarantees for both function estimation and variable selection. We design a Markov chain Monte Carlo algorithm for posterior computation. We evaluate the method's empirical performances through synthetic experiments and demonstrate its practical utility through real world applications.