Choosing of the best estimate of the parameters of the multiple linear regression model of infertility using the weighted least squares (WLS) and Robust M

You are currently viewing Choosing of the best estimate of the parameters of the multiple linear regression model of infertility using the weighted least squares (WLS) and Robust M

Choosing of the best estimate of the parameters of the multiple linear regression model of infertility using the weighted least squares (WLS) and Robust M

Choosing of the best estimate of the parameters of the multiple linear regression model of infertility using the weighted least squares (WLS) and Robust M

Saif Hussein Ali
Ahmed Hamza Abood 
Shrook.A.S.AL-Sabbah
University of Karbala \ collage of Administration and economics statistics department


Abstract:
In this research, methods were used to estimate the parameters of the multiple regression model, namely the method of the weighted least squares (WLS) and the robust- M methods, to show the effect of the presence of outliers values ​​in the observation values ​​on the regression model. A simple random sample of 151 infertile individuals are drawn as dependent variable (size of egg), and Independent variables (age, weight, infertility type, LH Hormone, FSH Hormone, PL Hormone) to compare these methods and to determinate the best method of estimation using measures mean square error MSE  and Determination Coefficient , and through the program Stata we found that the best method to estimate was robust- M (Turkey’s biweight)