A
AcadiFi
QD
QuantFinance_Dev2026-04-01
cfaLevel IIQuantitative MethodsMachine Learning

What is regularization in machine learning, and why should CFA candidates care about LASSO vs. Ridge?

The new Quantitative Methods section for CFA Level II now covers machine learning topics including regularization. I have a traditional finance background and I'm not sure why we need to learn about LASSO and Ridge regression. How do these techniques improve on ordinary least squares, and can someone give a finance-relevant example of when you'd use each?

156 upvotes
Verified ExpertVerified Expert
AcadiFi Certified Professional
Regularization adds a penalty term to OLS regression that discourages large coefficient values, addressing overfitting and multicollinearity — two critical problems in financial modeling. Ridge regression shrinks coefficients toward zero, while LASSO can force them to exactly zero, performing automatic variable selection.

Sign up to read the full expert answer

Get access to detailed explanations, worked examples, and expert insights.

📊

Master Level II with our CFA Course

107 lessons · 200+ hours· Expert instruction

#regularization#lasso#ridge-regression#machine-learning#overfitting#cross-validation