ML people: in the nn-nlp primer I say that the loss function is always positive or zero, and a reviewer questions this assertion. >
-
@fchollet can you give an example of negative loss?
@yoavgo any quantity you are trying to maximize, e.g. cosine proximity
-
-
@yoavgo makes more sense to take the opposite of the quantity than change the sign of the gradient updatesThanks. 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.