[felipellrocha]

Yep. I would like to see an article that implements everything without using matrices first, then creates the matrices library with you, and refactors everything over.

So much learning that we're missing by not going through this step.

[cr0sh]

A good course that comes close to this would be the Coursera Machine Learning course (what used to be known as "ML Class" by Andrew Ng).

It uses Octave - but you first do everything (in the section on NN) "by hand" - building and looping for the matrix operations. Only after you've gone that far, does he (Ng) introduce the fact that Octave has vector/matrix primitives...

I took the original ML Class in the Fall of 2011; it was a great class, and opened my eyes a great deal on the topic of machine learning and neural networks, which I had struggled with understanding in the past (mainly on what and how backprop worked).

[fartcannon]

To me it feels a bit like that joke about drawing instructions. "1. Draw some circles. 2. Now draw the rest of the owl."