Improved log-Gaussian approximation for over-dispersed Poisson regression: application to spatial analysis of COVID-19

TL;DR

This paper introduces a fast, stable, and accurate log-Gaussian approximation method for over-dispersed Poisson regression, improving estimation accuracy especially with many zeros, and demonstrates its application to COVID-19 spatial analysis.

Contribution

It develops a closed-form inference method for over-dispersed Poisson models using mode-based log-Gaussian approximation, addressing stability and efficiency issues.

Findings

The method reduces estimation error compared to existing approaches.

It performs well with many zeros in count data.

Applied to COVID-19 data, it reveals changing influences over time.

Abstract

In the era of open data, Poisson and other count regression models are increasingly important. Still, conventional Poisson regression has remaining issues in terms of identifiability and computational efficiency. Especially, due to an identification problem, Poisson regression can be unstable for small samples with many zeros. Provided this, we develop a closed-form inference for an over-dispersed Poisson regression including Poisson additive mixed models. The approach is derived via mode-based log-Gaussian approximation. The resulting method is fast, practical, and free from the identification problem. Monte Carlo experiments demonstrate that the estimation error of the proposed method is a considerably smaller estimation error than the closed-form alternatives and as small as the usual Poisson regressions. For counts with many zeros, our approximation has better estimation accuracy than conventional Poisson regression. We obtained similar results in the case of Poisson additive mixed modeling considering spatial or group effects. The developed method was applied for analyzing COVID-19 data in Japan. This result suggests that influences of pedestrian density, age, and other factors on the number of cases change over periods.