Hi,
In the back propagation video how does Andrew get the dZ_[1] result as W_[2].T dZ_[2] * g_[1]'(Z_[1])?
He skips the derivative of da_[1] so I have no idea how to get dZ_[1].
Hi,
In the back propagation video how does Andrew get the dZ_[1] result as W_[2].T dZ_[2] * g_[1]'(Z_[1])?
He skips the derivative of da_[1] so I have no idea how to get dZ_[1].
The point is that he doesn’t explain any of this really. This course is designed not to require knowledge of calculus. That’s the good news. But of course that means the bad news is that you just have to accept the formulas as he gives them to you.
Here’s a thread that discusses this in more detail and points you to some other resources.
Okay i see. Appreciate it
And here’s another thread on this same subject.