a proposed conditional method for estimating arma(1, 1) model

This paper aims to study the parameters estimation methods of the stationary mixed model (autoregressive-moving average) of low order arma (1, 1) regarding to time domain analysis in univariate time series. using the approximating methods: back forecasting (bf), classical conditional maximum likelihood (cc) and proposed conditional maximum likelihood (pc). a comparison is done among the three methods by mean squared error (mse) using several simulation experiments; the obtained results from the empirical analysis indicate that the accuracy of the proposed conditional method is better than the classical conditional method.