A Bayesian Approach of Predicting the Movement of Internally Displaced Persons.

This paper proposes an approximate Bayesian model to predict the number of internally displaced people arriving to a location. Locations are characterized by their elevation, distance from point of departure, and land cover. The model is applied to the population and terrain data of the North Kivu province in the Democratic Republic of Congo (DRC). Results suggest that distance captures about 67% of the influence on the choice of destination; elevation captures 9%, and land cover 24%.