Bayes Method & Goal programming to estimate the regression parameters

A thesis Submitted to the council of the college of administration and Economic\ University of Karbala, as partial fulfillment of the requirements for the degree of Master of Science in statistic

By
Ahmed Turki Abd Ali
Supervised by
Asist-Prof. Dr. Jassim N. Hussain

Abstract:
Estimation of  the parameter of regression model still on important subject inspite of a lot of researches have been  written about it. These studies are differ a cording to the methods or Baysien.
We will compar in this thesis the classical methods which are represented by ordinary least squares method with the Bayesian method which are represented by the method depend on (NON- informative prior density function and normal conjugate prior function). In addition we camper onther  two method the first is goal programming and proposed method which depent on the percintial.
The results of empirical studie show that the estimation of the parameter is equv. lant  in both (OLS and Bayes NON ). Mean  while ,the the best method is Bayes NC method followed be the proposed method in other side, We choose aleta sutable to the analysis Hypotheses in the applied side and we all these method to estimate the parameters of the Regression model. Then we  get that the best method also is Bayes NC and follow by  OLS method.