Where did these derivations come from? From last week lectures, the following was concluded regarding derivation of dZ, dA
dZ = dL/dZ = A - Y
dA = dL/dA[2] = -Y/A + (1-Y)/(1-A)
Moreover, there’s no mention of the loss function L.
Where did these derivations come from? From last week lectures, the following was concluded regarding derivation of dZ, dA
dZ = dL/dZ = A - Y
dA = dL/dA[2] = -Y/A + (1-Y)/(1-A)
Moreover, there’s no mention of the loss function L.
@Musab_Bin_Gulfam unfortunately I don’t have the video right in front of me at the moment, but everything back propagates from the loss, thus ‘chain rule’ and why you don’t see it.
It is percolating down.
Hi,
Thanks for the response, here is the video.
Is there any derivation listed in some other thread? Or some useful resource, please? These derivations just went right on top of my head.
Here’s a thread from Mubsi and Eddy that covers those derivations.
@paulinpaloalto thanks for helping; I know my poetry is not great, but I did not want to call you out, but I was like, oh, Paul would know this.