Index error in the diagram of the chapter calculas for ML and data science,

Course Q&A Math for Machine Learning Calculus for Machine Learning and Data Science
,
1

Hi everyone,
I just wanted to point out a small error in the diagram at 0:41 in the chapter “Classification with a Neural Network – Minimizing Log Loss”.

In the standard notation, WijW_{ij}Wij​ means:

  • iii → destination neuron

  • jjj → source (current) node

However, in the diagram, these indices appear to be reversed, which made the weight directions unclear. Because of this, I initially had a hard time correctly deriving the partial derivative formulas during backpropagation.

I’m sharing this so others are aware of it and don’t get confused when working through the gradient derivations.

Hope this helps!

2

Thanks for your report.