Convolution_model_Step_by_Step_v1 Backward Pass formulas

Course Q&A Deep Learning Specialization Convolutional Neural Networks
1

Convolution_model_Step_by_Step_v1 There seems to be missing equations for backward pass. Code says refer to equations above but backward pass starts abruptly without any introduction.

Missing topics from table of contents:

5 - Backpropagation in Convolutional Neural Networks (OPTIONAL / UNGRADED)
5.1 - Convolutional Layer Backward Pass
5.1.1 - Computing dA
5.1.2 - Computing dW
5.1.3 - Computing db

Thanks in advance.

2

They don’t cover the intuition behind the back prop. But if you want to dive deep, you can by using Google/YouTube or books/articles.

3

Thank you for quick reply. I meant assignment itself has following comment:

                    # Update gradients for the window and the filter's parameters using the code formulas given above

But there are no formulas. Also name of file is v1. Seems like formulas and section was removed accidentally from this version.

Updated thread above about missing topics:
5 - Backpropagation in Convolutional Neural Networks (OPTIONAL / UNGRADED)

4

But I can see this in my notebook, v1.1. Maybe you deleted this mistakenly. Here is the guide on how to get a fresh copy of your assignment. It will undo all your code, so, it’s better to submit your graded code first.

Capture
Capture816×760 95.2 KB

5

Thank you. Dunno when I deleted them.

Best,
Rahul