You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
yes, but I think we should think about how the output should be. For example, we might
have flash(Y,biscale=TRUE) output L and F that correspond to the original Y (even thought they were computed first by biscaling). My hope is that we might be able to compare objective functions
based on a "biscale model" vs nonscaled model to decide if biscaling is a good idea.
we should think about possibility of dealing with biscaling (eg as in softImpute)
The text was updated successfully, but these errors were encountered: