-
Notifications
You must be signed in to change notification settings - Fork 4.3k
Loss Functions and Metrics
Frank Seide edited this page Aug 19, 2016
·
23 revisions
CNTK contains a number of common predefined loss functions. In addition, custom loss functions can be defined as BrainScript expressions.
Computes the cross entropy
CrossEntropyWithSoftmax (L, z)
-
L
CrossEntropyWithSoftmax_new (L, z, tag='') = Minus (ReduceLogSum (z), TransposeTimes (L, z), tag=tag)
ErrorPrediction (L, z)
ErrorPrediction_new (L, z, tag='') = Minus (BS.Constants.One, TransposeTimes (L, Hardmax (z)), tag=tag)