You may as well attempt heading in the opposite way: Discover the 'outliers' & see how they vary from the remainder of your data. Checking out the residuals vs equipped plot, I see that many the 'outliers' have predicted values in the center. W/o recognizing much more about y

Article Under Review