Sankhya: The Indian Journal of Statistics
1995, Volume 57, Series A, Pt. 2, pp. 287--298
ESTIMATION FOR STATIONARY AR(1) MODELS WITH NONCONSECUTIVELY OBSERVED SAMPLES
DONG WAN SHIN, Ewha Womans University
SAHADEB SARKAR, Oklahoma State University
SUMMARY. For the stationary first order autoregressive models with irregularly observed or missing data, it is shown that the maximum likelihood estimator is consistent and asymptotically normal. Simulation results are presented on the relative performance of the maximum likelihood estimator of the autoregressive parameter. We discuss a simple estimator which performs very well under various conditions, and which when used as an initial value for computing the one-step Newton-Raphson estimator of Reinsel and Wineck (1987) results in excellent empirical efficiency of the latter, relative to the maximum likelihood estimator.
AMS (1991) subject classification. 62M10, 62F12.
Key words and phrases. Asymptotic normality, autoregressive model, consistency, maximum likelihood estimator, missing or unequally spaced data.
Full paper (PDF)
This article in mathematical reviews.