Disease Mapping and Regression with Count Data in the Presence of Overdispersion and Spatial Autocorrelation: A Bayesian Model Averaging Approach

摘要: This paper applies the generalised linear model for modelling geographical variation to esophageal cancer incidence data in Caspian region of Iran. The have a complex and hierarchical structure that makes them suitable analysis using Bayesian techniques, but with care required deal problems arising from counts events observed small areas when overdispersion residual spatial autocorrelation are present. These considerations lead nine regression models derived three probability distributions count data: Poisson, Poisson negative binomial, different structures. We employ framework variable selection Gibbs sampling based technique identify significant risk factors. deals situations where number possible on combinations candidate explanatory variables is large enough such calculation posterior probabilities all difficult or infeasible. evidence applying methodology suggests strategies use binomial work well provide robust basis inference.