Time varying coefficient AR and VAR models

From MaRDI portal





Analysis of time varying coefficients AR and VAR models is surveyed in this contribution (previous work by Kitagawa, Jing and other authors). The procedure for estimating these models consists in introducing the state space model based on Bayesian smoothness priors. The estimates of the time varying coefficients and the likelihood of the hyperparameters of the models are obtained by the Kalman filter. The hyperparameters are estimated by the method of maximum likelihood. The AR order is determined by the minimum AIC procedure. The estimation of the time varying spectrum and cross-spectrum is discussed. An example of seismic data analysis using the time varying coefficients VAR model is given. For the entire collection see [Zbl 0907.00020].











This page was built for publication: Time varying coefficient AR and VAR models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4249441)