Estimating the Coefficients of Polynomial Regression Model when the Error Distributed Generalized Logistic with Three Parameters
1Ahmed Dheyab Ahmed, Ebtisam Karim Abdulah, Baydaa Ismael Abdulwahhab
In this research, estimates of polynomial regression model coefficients are derived when the random error is logistically distributed with three parameters representing scale, shape and location. By employing simulation method, the model coefficients are compared using methods of General Least Squares and the modified maximum likelihood. Based on statistical criteria, the Mean Square Error (MSE) indicated that the modified maximum likelihood (MML1) method is the best. This method is applied to data of Gross Domestic Product (GDP) at current prices in Iraq.
Polynomial regression, Generalized Logistic distribution, Modified maximum likelihood method, General least squares, Mean square error.