On the asymptotic normality of posterior distributions in the multiparameter case.
Bayesian Statistics 4
825 - 835.
Current approaches to demonstrating asymptotic posterior normality in the multiparameter case fail to cover certain cases where the data arise from a stochastic process. In this paper we present a set of hypotheses for asymptotic posterior normality which are satisfied in such cases. Continuity-type conditions on observed information are imposed on suitable shrinking neighbourhoods and workable conditions relating to the tail behaviour of the posterior distribution are given. The theory is illustrated with an example of a two-parameter nonhomogeneous Poisson process.
|Title:||On the asymptotic normality of posterior distributions in the multiparameter case|
|Keywords:||Asymptotic posterior normality, Inference for stochastic processes, Nonhomogeneous Poisson process|
|UCL classification:||UCL > School of BEAMS > Faculty of Maths and Physical Sciences > Statistical Science|
Archive Staff Only