@fchollet Does that include held out data they actually evaluated the leaderboard on? Did they ever even publish that?
Beating the winners of the 2009 Netflix prize ($1M) with a super simple Deep Learning solution, 15 lines of Keras:https://karthkk.wordpress.com/2016/03/22/deep-learning-solution-for-netflix-prize/ …
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
@fchollet not on the same test set though? -
@syhw Right. Not sure if the "official" test set is available. But 2% difference sounds like it would be significant regardless. - Show replies
New conversation -
-
-
@fchollet iirc netflix prize used a custom train/test split intentionally to make it harder -
New conversation -
-
-
@fchollet The author used a different data set for the score. Against the same validation data set, the score was much worse at 0.9357.Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.