The spatial information contained in the epidemiological data has huge value for the study of epidemics. First, this dissertation introduced the main theories of spatial statistics, including the common methods of analysis and regression models. Second, for different types of spatial data, including single scale spatial and spatio-temporal data, multiscale spatial and spatio-temporal data, and multivariate disease data, we applied different regression models to analyze them.By comparing the DIC values, we get the relatively optimal model to smooth the regional relative risk and solve the problem of sparsely populated areas generating extreme values of the relative risk. |