A Note on Estimation for the First Order Autoregressive Process | ||||
The Egyptian Statistical Journal | ||||
Volume 31, Issue 1, June 1987, Page 62-76 PDF (9.41 MB) | ||||
Document Type: Original Article | ||||
DOI: 10.21608/esju.1987.428901 | ||||
![]() | ||||
Authors | ||||
A.M. Abouammoh; A.A. Abd-Alla | ||||
King Saud University Riyadh, Saudi Arabia. | ||||
Abstract | ||||
The asymptotic estimation of the parameters of the first order stationary autoregressive process with zero mean is investigated. The exact and the asymptotic Bayes esti- mates of the weight and the error variance parameters are derived for different priors. The inverse of the error variance is assumed to be distributed priori according to gamma distribution. The weight parameter is given a U-shaped, uniform and bell shaped prior distributions. A numerical study of various estimation of parameters with the maximum likelihood estimates is shown for large sample sizes. It was found that for large sample size the Bayes estimates of the weight parameter tends to conditional maximum likelihood in case of uniform or U-shaped prior | ||||
Keywords | ||||
Stationary process; Maximu, Likelihood estimation; Bayes estimation | ||||
Statistics Article View: 81 PDF Download: 23 |
||||