Non-centred parameterisations for hierarchical models and data augmentation.

Abstract

SUMMARY In this paper, we will compare centered and non-centered parameterisations for classes of hierarchical models. Our examples will include variance component models, random effect models, hidden Markov process models, and partially observed diffusion models. We will investigate the construction of non-centered methods by the use of state space expansion techniques, and will introduce methods for devising partially non-centered parameterisations, many of which are data-dependent.