QMRITools`
QMRITools`
TverskyLossLayer
represents a net layer that computes the Tversky loss by comparing input class probability vectors with the target class vector.
does the same but b defines the tversky beta factor. With beta = 0.5 its is the Dice coefficient. Here alpha + beta = 1.
Examples
In[1]:=1

✖
https://wolfram.com/xid/0l27f6k4bj3ugzr4mox9jllskx6w1p095e-ipw808