الفهرس 
Title : Statistical inference of the parameters in autoregressive models
AbstractOnly 14 pages are availabe for public view
 |  | from | 180 |  |  |
| | | from | 180 | | |
Abstract
This thesis aims to introduce the unconditional autoregressive models which had a little attention in theoretical econometric literature, which was the first introduced by Box and Jenkins (1976). In this thesis, the previous studies concerning the estimation methods of AR models will be reviewed and the methods of estimation for the linear models and for the Autoregressive models will be introduced. Furthermore, the different method of AR (1) model without constant will be discussed. In addition, the properties (linearity, biasness and consistency) of the estimators are investigated theoretically. On the other hand, the (ML) method was introduced to estimate the unknown parameters of AR (2) model without constant. Finally, a simulation study has been designed to compare different methods of estimation (UML, UWLS and UWS) for different sample sizes for AR (1) model without constant term, based on the bias and RMSE criteria’s. This simulation study is evaluated for several values of the parameters and different values of the variance and mean error used