Skip to content
Free public beta until December 4, 2026: every course is open, no card needed. After that, the first chapter of each course stays free and full access is a one-time $39 ($19 for up to 100 founding members chosen for helpful feedback). Your progress is kept. How the beta works

Regularization

Adding a penalty term to a model's loss to discourage overly large or complex fits.

A modification to a model's training objective that adds a penalty on coefficient size (or model complexity), trading some training-set fit for reduced variance and better generalization -- ridge (L2), lasso (L1), and elastic net (a blend) are the three linear-model instances this course covers.

Examples

  • Ridge regression's lambda term penalizes the sum of squared coefficients.

Read the full section →
Graph settings

Calculator