FIML estimation of dynamic econometric systems from inconsistent data

From MaRDI portal





This paper deals with a Bayesian and generalized least-squares procedure suitable for econometric estimation of discrete-time dynamic models on the basis of data displaying inconsistencies, corrupted by measurement errors, related to variables which cannot be observed directly, and may therefore involve subjective and conflicting views. The paper shows how such noisy data may be used to identify econometric systems by FIML estimation and discusses special cases where the Kalman- Bucy filter can be avoided and when the log-likelihood criterion is feasible. Numerical algorithms, extensions and applications as well are discussed.



Cites work









This page was built for publication: FIML estimation of dynamic econometric systems from inconsistent data

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3675389)