I'm working on it with weighted regularization. The problem is that OWLQN doesn't work nicely with Updater now since all the L1 logic should be in OWLQN instead of L1Updater.
Sincerely, DB Tsai ------------------------------------------------------- My Blog: https://www.dbtsai.com LinkedIn: https://www.linkedin.com/in/dbtsai On Fri, Jul 18, 2014 at 1:38 PM, Debasish Das <debasish.da...@gmail.com> wrote: > Hi, > > I thought OWLQN is already merged to mllib optimization but I don't see it > in the master yet... > > Are there any issues in merging it in ? I see there are some merge > conflicts right now... > > https://github.com/apache/spark/pull/840/ > > Thanks. > Deb