A Comparison of the Estimation Methods for the Non-Linear Parameter of First Order Moving Average Model

Abstract

A lot of researchers face a difficulty in estimating parameter of moving average model of non-linear parameter. For that reason, we can't use the traditional methods of estimation.This research aims to compare four methods in estimating parameter of first order moving average model by using mean absolute percentage error as a criterion comparison with consideration (sample sizes, different initial value parameter, and different distributions of random errors of the sample ),The second method has shown to be the best among other methods of sample sizes, except where the initial value parameter of the model is 0.13 and here the third method can be the best for geometric distribution and for all sample sizes.