When using coordinate descent for solving a lasso regression, does normalizing the features impact the convergence rate?
$\begingroup$
$\endgroup$
Add a comment
|