Bayesian Methods for Data Correlated in Time and Space

Abstract :

Previously difficult to fit and perform inference on, Bayesian models (and the programming and analytical frameworks to compute them efficiently) are enjoying a splendid renaissance across the board. Harnessing the novel Integrated Nested Laplacian Approximation (INLA) framework I carry out two investigations into the capabilities of Bayesian statistics to describe data with strong dependencies in time and space, shedding light on tough, but extremely rigurous, modelling techniques.

Download here