Bayesian forecasting' is a time series method of forecasting which (in the United Kingdom) has become synonymous with the state space formulation of Harrison and Stevens (1976). The approach is distinct from other time series methods in that it envisages changes in model structure. A disjoint class of models is chosen to encompass the changes. Each data point is retrospectively evaluated (using Bayes theorem) to judge which of the models held. Forecasts are then derived conditional on an assumed model holding true. The final forecasts are weighted sums of these conditional forecasts.
Few empirical evaluations have been carried out. This paper reports a large scale comparison of time series forecasting methods including the Bayesian. The approach is two fold: a simulation study to examine parameter sensitivity and an empirical study which contrasts Bayesian with other time series methods. KEY WORDS Bayesian forecasting Comparative methods-time series Exponential smoothing Stationarity Data-simulation Estimation-Kalman filter THE BAYESIAN FORECASTING MODEL Bayesian methods have frequently been used in forecasting applications, typically as a means of estimating the parameters of a model (e.g. Zellner, 1971). This approach adds little new to conventional methods of model estimation. However, Zellner also described an important new way of looking at the problem of deciding which of a number of distinct models best encapsulated both the researcher's prior views on each model's validity and the experimental evidence that had been collected. Both uses of Bayesian thinking were effectively incorporated in a method that has become known as Bayesian forecasting. It was proposed by Harrison and Stevens in their paper and its mathematical basis was made more explicit in a subsequent paper (Harrison and Stevens, 1971, 1976). While it was primarily a univariate time series method of forecasting, it was easily extended as the 1976 paper showed. Relatively few applications have been published, despite
The Gunning Fog Index for this paper is 14.