A Latent Gaussian Mixture Model for Clustering Longitudinal Data

TL;DR

This paper introduces a novel latent Gaussian mixture model tailored for clustering high-dimensional longitudinal data with many time points, combining mixture models with factor analysis techniques.

Contribution

It develops an extended mixture of common factor analyzers model specifically for longitudinal data and employs a modified EM algorithm for efficient parameter estimation.

Findings

Effective clustering demonstrated on real and simulated datasets.

Model selection guided by Bayesian information criterion.

Addresses gap in clustering methods for high-dimensional longitudinal data.

Abstract

Finite mixture models have become a popular tool for clustering. Amongst other uses, they have been applied for clustering longitudinal data and clustering high-dimensional data. In the latter case, a latent Gaussian mixture model is sometimes used. Although there has been much work on clustering using latent variables and on clustering longitudinal data, respectively, there has been a paucity of work that combines these features. An approach is developed for clustering longitudinal data with many time points based on an extension of the mixture of common factor analyzers model. A variation of the expectation-maximization algorithm is used for parameter estimation and the Bayesian information criterion is used for model selection. The approach is illustrated using real and simulated data.