4.5 Article

Dynamic ICAR Spatiotemporal Factor Models

Journal

SPATIAL STATISTICS
Volume 56, Issue -, Pages -

Publisher

ELSEVIER SCI LTD
DOI: 10.1016/j.spasta.2023.100763

Keywords

Areal data; Bayesian dynamic models; COVID pandemic; Dynamic linear models; Factor models; Intrinsic conditional autoregression

Ask authors/readers for more resources

We propose a novel class of dynamic factor models for spatiotemporal areal data, achieving dimension reduction by assuming that the process can be represented by a few latent factors. Each column of the factor loading matrix follows an intrinsic conditional autoregressive process to account for spatial dependence. We present two case studies, demonstrating the identifiability of our models and the utility of the framework in analyzing the drug overdose epidemic in the United States.
We propose a novel class of dynamic factor models for spa-tiotemporal areal data. This novel class of models assumes that the spatiotemporal process may be represented by some few latent factors that evolve through time according to dynamic linear models. As the dimension of the vector of latent factors is typically much smaller than the number of subregions, our proposed class of models may achieve substantial dimension re-duction. At each time point, the vector of observations is linearly related to the vector of latent factors through a matrix of factor loadings. Each column of this matrix may be seen as a vectorized map of factor loadings relating one latent factor to the vector of observations. Thus, to account for spatial dependence, we assume that each column of the matrix of factor loadings follows an intrinsic conditional autoregressive (ICAR) process. Hence, we call our class of models the Dynamic ICAR Spatiotemporal Factor Models (DIFM). We develop a Gibbs sampler for exploration of the posterior distribution. In addition, we develop model selec-tion through a Laplace-Metropolis estimator of the predictive density. We present two case studies. The first case study, which is for simulated data, demonstrates that our DIFMs are identi-fiable and that our proposed inferential procedure works well at recovering the underlying data generating process. Finally, the second case study demonstrates the utility and flexibility of our DIFM framework with an application to the drug overdose epidemic in the United States from 2015 to 2021.& COPY; 2023 Elsevier B.V. All rights reserved.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.5
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available