Training a Neural Network with back propagation

Course Q&A Machine Learning Specialization Advanced Learning Algorithms
1

I was really looking forward to an explanation for how a neural network is trained via back-propagation, but we really only covered which functions to call in Tensorflow. Is there an article//video someone could recommend that covers an explanation for what is actually happening during back-propagation? It’s odd to me that this wasn’t really covered since it’s a critical detail in understanding a neural network.

2

Hi!

I personally like this article here:
https://colah.github.io/posts/2015-08-Backprop/

1 Like
3

Yeah exactly! Why was this ignored completely? We should have had at least an Optional section providing an intuitive understanding of how back propagation works.

Hope a mentor can explain the rationale for skipping this topic entirely.

4

Hello @Prax,

I think there should be some ongoing effort about back propagation. Please check this message by a member of the course team, and there you can find a link to another thread which has some relevant materials.

Cheers,
Raymond

1 Like
5

Thank you @rmwkwok! I found that thread after commenting here. Looking forward to the upcoming video! :+1:

6

You are welcome @Prax!