Robust estimation of the seasonal autocorrelation of the par(1)‎ model

Abstract EN

In this article we are interested in the robust estimation of seasonal autocorrelation for the periodic autoregressive model of order one (PAR (1)).

We used three estimators for the first–lag seasonal autocorrelation including the classical moment estimator beside two new proposed robust estimators.

The effect of a single additive outlier contaminated in the time series is examined via bias and MSE.

We have also studied the effects of some other factors on the quality of those estimators.

The investigation is carried out using Monte–Carlo simulation.

The results show that our proposed estimators are robust whereas the moment estimator is not.

This conclusion is also assured via the bias and MSE of those estimators.