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.
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.