medium10 pts · first answer countsmachine-learningregularisation

When elastic net beats lasso

Your data has groups of strongly correlated features (for example several overlapping income measures). Why might elastic net be preferred over pure lasso?

A.Lasso keeps one of a group at random; elastic net shares weightB.Elastic net trains with no hyperparameters to tune at allC.Lasso cannot be used once features have been scaledD.Elastic net always produces more zero coefficients than lasso

Discussion 0

Join free to post in the discussion.

When elastic net beats lasso

Discussion 0

Related questions