الفهرس Only 14 pages are availabe for public view
 |  | from | 147 |  |  |
| | | from | 147 | | |
Abstract
The main objectives of this thesis are: Derive the estimators of the unrestricted SUR model under subset of regressors using a new approach, derive the first and second moment matrix of the estimators. Derive the estimators of the unrestricted SUR model under the orthogonal condition and derive the variance covariance matrix of the estimators. On the other hand, introduce the approximation propertics of seemingly unrelated unrestricted regression equations model, derive the bised estimators to order 0,({u023E}⁻¹) and derive covariance matrix of the feasible (SURE) estimator to order 0,({u023E}⁻²). Finally, derive mixed estimation of seemingly unrelated regression equation model and some finite sample properties results that discuss the model and their assumpyion. Explore the different types of information model that we have the sample information model and stochastic linear information model, unifying the two models in a single model this is called the feasible mixed generalized least square model FMGLS and derive the mixed model estimators. Derive the bias to 0,({u023E}⁻¹) and moment matrix to order 0,({u023E}⁻²) of feasible mixed seemingly unrelated unrestricted regression equation model