On Forecasting with Univariate Autoregressive Processes: A Bayesian Approach.

摘要

Using a normal-gamma prior density for the parameters of a p-th order autoregressive process, the Bayesian predictive density of k future observations is derived. It is shown that the joint predictive density of k future observations may be expressed as the product of k univariate t densities. Our results are illustrated with one-step ahead forecasts employing an AR(1) model with a conjugate prior density for the parameters. (Author)