In a previous post, we demonstrated that ridge regression (a form of regularized linear regression that attempts to shrink the beta coefficients toward zero) can be super-effective at combating overfitting and lead to a greatly more generalizable model. This approach… Continue reading →