Towards joint disease mapping using penalised complexity priors

Abstract

This thesis goal is to assess whether penalized complexity priors can be used successfully in a joint disease modeling framework and have similar properties as when used for modeling only one disease. Penalized complexity priors have never been used in models considering more than one disease at once and dealing with this is the core of this thesis. This thesis also has a didactic purpose, and is made so that the reader sees the progress that have been made throughout the ten months this thesis took to complete. This thesis shows that under certain simplifying assumptions, and with an educated choice of model that penalized complexity priors provide encouraging results when jointly modeling two diseases. Further work would be required to assess the exact extent of these results.