A novel method for estimating the parameter of a Gaussian AR(1) process with additive outliers release_7mqo5d3oxzgadokbt7f6yr3xzm

Abstract

A novel estimator for a Gaussian first-order autoregressive [AR(1)] process with additive outliers is presented. A recursive median adjustment based on an -trimmed mean was applied to the weighted symmetric estimator. The following estimators were considered: the weighted symmetric estimator (ˆ W ), the recursive-mean-adjusted weighted symmetric estimator (ˆ R W  ), the recursive-median-adjusted weighted symmetric estimator (ˆ Rmd W  ), and the weighted symmetric estimator using adjusted recursive median based on the -trimmed mean (ˆ Tm Rmd W   ). Using Monte Carlo simulations, the mean square errors (MSE) of the estimators were compared. Simulation results showed that the proposed estimator, ˆ Tm Rmd W    , provided a smaller MSE than those fromˆWfromˆfromˆW  , ˆ R W   andˆRmdandˆandˆRmd W   for almost all situations.
In text/plain format

Archived Files and Locations