Using Ridge Regression to Estimate Factors Affecting the Number of Births. A Comparative Study

Abstract

Ridge regression method is a biased regression analysis method that is used when the data suffer from multicollinearity problem between its explanatory variables, as the use of the least-squares method in analyzing such data will lead to incorrect estimates of the parameters of the regression coefficients and thus lead to incorrect predictions. For this reason, many methods for obtaining the ridge parameters in ridge regression were used previously in dealing with the problem of multicollinearity data, to overcome this problem, we used our formula in previous research named, the performance of the new ridge regression parameter and we applied it to a real data concerning factors affecting the number of births on a group of women who visited the health centers in Babel Governorate, as it was found that these data suffer from collinearity (multicollinearity problem). A comparison was made between this method and the other methods used previously, and results showed the effectiveness of this method as it gave better results than the methods used previously.