I want to implement back-propagation neural network algorithm in raw code, not using any library as my interest to understand in depth operations/customization of this algorithm.

In this regard, I need to follow step-by-step simulation of back-propagation algorithm. More precisely, what code I should write in R language to each of steps in back-propagation neural network. Can anyone show me the close explanation of steps in this algorithm.

I don't need any code. Just show me what operations is done in each steps.

FYI: for reviewers, this is not my assignment. I'm doing my assignments using neuralnet library. Please leave a comment before any report.

  • $\begingroup$ Would a step by step derivation of a vanilla neural network help you? $\endgroup$ Commented Dec 15, 2015 at 20:01
  • $\begingroup$ @ArmenAghajanyan Can you provide any example? $\endgroup$
    – mmr
    Commented Dec 16, 2015 at 3:58

1 Answer 1


If you just want to see how the internals work, the best resource for a beginner, that I have found, is this series of videos by Welch Labs. It goes through each step of a basic feed-forward backpropogation model including the calculus steps so you understand what is happening during each step. It is all done in Python but you can definitely convert it to R if that is your goal. That would be a great learning exercise.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.