matteocourthoud/Machine-Learning-for-Economic-Analysis

Missing Scaling in Lasso CV?

Jan-Niklas-Doerr opened this issue · 0 comments

Hi,

First, I appreciate your work and sharing your knowledge; it is beneficial. However, why did you decide not to scale the X in the 10-fold CV to determine the best lambda for Lasso (05_regularization In [57])? If you do, you get a different result, i.e., that it is best to not use the penalty term at all.