Solved – PCA vs ridge regression for multicollinearity

Question

Orthogonalization via PCA and ridge regression are two common methods to account for multicollinearity for linear regression models. When would you use one over the other?

Answer 1

When the cross-validated error of one method is lower than the other. I would also look into lasso regression and elastic net regression.