How can I implement this?

How to implement J(+) = J(Theta+)

You just wrote the forward propagation function, right? Their notation is a little strange here, since normally what we mean by J is cost, but they seem to be using it to represent \hat{y} here. That means the output of forward propagation.

I’d venture that this assignment needs a bit of an update.

Since the gradients are the partial derivative of the cost, it’s not cool to use the partial derivative of y_hat instead.

1 Like