C2W1 Gradient Checking Section 5 - Minor Detail

Hi there DeepLearning.AI crew,

This is just a minor thing I noticed, but wanted to float up. At the start of Section 5, a couple of cells define forwardprop and backprop functions, but then the following markdown mentions obtaining “some results on the fraud detection test set” when there is not a cell that does any such thing. Doesn’t affect the exercises, but wanted to point it out!

Yes, perhaps this is a little “hokey”, but they are constructing an imaginary scenario in which you might want to use Gradient Checking. It’s just referring back to that theme as stated in this earlier section:

2 - Problem Statement

You are part of a team working to make mobile payments available globally, and are asked to build a deep learning model to detect fraud–whenever someone makes a payment, you want to see if the payment might be fraudulent, such as if the user’s account has been taken over by a hacker.

You already know that backpropagation is quite challenging to implement, and sometimes has bugs. Because this is a mission-critical application, your company’s CEO wants to be really certain that your implementation of backpropagation is correct. Your CEO says, “Give me proof that your backpropagation is actually working!” To give this reassurance, you are going to use “gradient checking.”

Let’s do it!

Hi Paul! Haha, okay, so it isn’t intended to be taken literally, but as an “imagine we obtained some results that make us doubt our model” kind of thing. That seems reasonable. I definitely expected to have some canned code to run to set up the statement when I came across it though.