Regularized Regression: Ridge, Lasso, and Elastic Net

What you'll build

• Ridge, lasso, and elastic net models fit and compared on the same data
• A workflow for tuning the regularization strength with cross-validation
• A reusable function that picks the model that generalizes best

What you'll learn

• What regularization does and why it beats plain regression on real data
• The difference between ridge, lasso, and elastic net, in code and intuition
• How lasso performs feature selection for you
• How to choose the penalty with cross-validation instead of by hand

About Greg

Associate Professor and Director of the Center for Data Science and Consulting at Loyola University Chicago, with research spanning statistical disclosure control, statistical genetics, and statistics in sports.