تقدير معلمات أنموذج الانحدار الخطي المتعدد في ظل وجود مشكلة التعدد الخطي

Abstract EN

In order to get at a model that leads to accurate predictions, it is necessary to search for the method of selecting the most important variables to be included in the model, especially when the data under study suffer from the problem of Co-Linearity, and the presence of many phenomena in our daily life, The main objective in this paper to compare among the methods of estimate unbiase parameters of the regression model, OLS, Unbiase Ridge Regression and Lasso Regression and we consider MSE and P-Value as measures of comparison, a random sample of 100 women's fertility was withdrawn to study the factors affecting the number of children born as dependent variable and many independent random variables is (woman age, the age at married, graduate of woman, graduate of husband, women's weight, women's use of contraception, women's smoking, Age of Husband, job of husband, Marriage Period, Number of Dead Children, Hours of Exercise per Week, hours of female sleep per day, women taking drugs, duration of breast feeding, mother's job), data were suffering from a linear multiplicity problem, It was found that the Lasso regression method has the best estimation methods for owning the MSE, followed by the method of the unbiased ridge