Advances in Spatial Integer-Valued Time Series Modeling

This study develops an empirical Bayes (EB) spatial hurdle INGARCH model for weekly dengue fever counts and compares it with a spatial zero-inflated generalized Poisson (ZIGP) INGARCH framework, both capturing spatio-temporal dependence and excess zeros. The EB-hurdle model introduces a data-adaptive prior that reduces model complexity and enhances parsimony and stability while retaining flexibility for dynamic zero inflation. In spatial INGARCH models, the zero-generating mechanism depends on lagged outcomes, implying that covariate effects influence the intensity equation indirectly. To enhance epidemiological relevance, seasonal patterns are incorporated into the log-intensity equations using Fourier-based harmonic terms and meteorological covariates. Model inference is conducted within a Bayesian framework. The results highlight the distinct roles of seasonal and environmental drivers in dengue transmission and demonstrate that Fourier-based periodic components provide an effective alternative when meteorological data are limited or unavailable. Overall, the empirical Bayes approach offers a parsimonious and stable improvement over conventional hurdle INGARCH models.

Bayesian inference

Dengue fever

Fourier series

INGARCH models

Spatio-temporal modeling

Zero-inflated count data 

International Society for Bayesian Analysis (ISBA)