A RANDOM PARAMETER PROCESS FOR MODELING AND FORECASTING TIME SERIES

From MaRDI portal
Publication:3716154





A generalized autoregressive (GAR) process \(\{\) Z(t); \(t=0,\pm 1,...\}\) is defined to satisfy the recurrence relation \(Z(t)=A_{\theta}(t)Z(t- 1)+u(t)\), where \(\{A_{\theta}(t)\); \(t=0,\pm 1,...\}\) is itself a stochastic process depending on a vector parameter \(\theta\) and where \(\{\) u(t); \(t=0,\pm 1,...\}\) is white noise with \(Eu^ 2(t)=\sigma^ 2.\) This paper develops theory and methodology and implementing the class of GAR processes for time series modeling and forecasting. Conditions on the parameter process \(\{A_{\theta}(t)\); \(t=0,\pm 1,...\}\) are obtained for the existence of a GAR process; necessary and sufficient conditions on \(\{A_{\theta}(t)\); \(t=0,\pm 1,...\}\) for existence of a stationary GAR process are also obtained. Procedures are developed for computing maximum likelihood estimates of the parameters \(\theta\) and \(\sigma^ 2\) and for computing the minimum mean squared error forecasts for GAR processes.




Cited in
(20)








This page was built for publication: A RANDOM PARAMETER PROCESS FOR MODELING AND FORECASTING TIME SERIES

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3716154)