Replies: 2 comments 1 reply
-
It is currently not possible in scikit-learn, see #11566. |
Beta Was this translation helpful? Give feedback.
0 replies
-
This would be a very useful feature to have. It looks like #11566 is kind of stale. Is there any progress towards this? Or perhaps a useful workaround that others have implemented? Thanks! |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I am running a Ridge model and from the docs I see that the
alpha
parameter is either a scalar or length ofn_targets
. Since the alpha appears in a sum of the magnitude coefficients wouldn't it make more sense to be able to pass alpha as a vector of the size n_predictors - so as to be able to apply different regularization parameters to the coefficients of different predictors? Is this an option in any of the models involving regularization?Beta Was this translation helpful? Give feedback.
All reactions