ABSTRACT
The analysis found for spatial data can be extended into the time domain without significant difficulty. The intensity specification can be used as a basis for the development of Bayesian models for case events. If it can be assumed that the events form a modulated Poisson process in space-time, then a likelihood can be specified, as in the spatial case. The parameter estimates suggest that the overall rate and space-time component are well estimated but the spatial and temporal effects. More recent examples of spatio-temporal modeling include extensions of mixture models, which examine time periods separately without interaction, and with the use of a variant of a full multivariate normal spatial prior distribution for the spatial random effects, and the extension of the Knorr-Held and Besag model to include different forms of random interaction terms.
