Bayesian approach to autoregressive processes.


Modeling hydrologic processes. Proc. Fort Collins 3rd international hydrologic symposium, July 1978, pp.165-180, 1979 (Scopus) identifier


A general approach leading to the Bayes estimate of any parameter of a linear stationary process has been presented and necessary analytical expressions have been derived. In particular, the application has been performed for the ARIMA (1, 0, 1) process which subsumes most of the autoregressive processes employed in hydrology. Especially, the posterior probability distribution function of phi which is the controlling parameter of a long-term behaviour of the ARIMA (1, 0, 1) process has been derived. - from Author