Tut 6 Flashcards
1
Q
Show how using E[-log(D(G(z))] helps with vanishing gradient in GANs
A
Using this cost function instead, we find
KCL Pattern Recognition, Neural Networks And Deep Learning > Tut 6 > Flashcards
Show how using E[-log(D(G(z))] helps with vanishing gradient in GANs
Using this cost function instead, we find