Multicollinearity Detection And Regression Analysis Of Intensity Of Stroke In Patients

Abstract

In this paper, data set regarding stroke diagnosis and its variables are considered. A regression equation is found with Barthel index as a response variable and the other variables as continuous predictors. The same was analyzed for multicollinearity by calculating the variance inflation factor (VIF). The variables with high value of VIF were removed to give a better regression equation devoid of multicollinearity. The correlation matrix is calculated before and after removing multicollinear variables. The results showed the significant improvement in regression analysis after removing multicollinear variables