Scalable Bayesian modeling for smoothing disease risks in large spatial data sets

TL;DR

This paper introduces a scalable Bayesian approach for smoothing disease risks in large spatial datasets, effectively handling high-dimensional areal data with improved computational efficiency.

Contribution

It presents a novel divide-and-conquer Bayesian method implemented in R for high-dimensional spatial risk smoothing, addressing a gap in existing methods.

Findings

Method outperforms standard approaches in goodness of fit.

Significantly reduces computational time.

Effective in high-dimensional areal data analysis.

Abstract

Several methods have been proposed in the spatial statistics literature for the analysis of big data sets in continuous domains. However, new methods for analyzing high-dimensional areal data are still scarce. Here, we propose a scalable Bayesian modeling approach for smoothing mortality (or incidence) risks in high-dimensional data, that is, when the number of small areas is very large. The method is implemented in the R add-on package bigDM. Model fitting and inference is based on the idea of "divide and conquer" and use integrated nested Laplace approximations and numerical integration. We analyze the proposal's empirical performance in a comprehensive simulation study that consider two model-free settings. Finally, the methodology is applied to analyze male colorectal cancer mortality in Spanish municipalities showing its benefits with regard to the standard approach in terms of goodness of fit and computational time.