So I'm currently reading about neuroevolution (NEAT, WANN) and trying to make my own implementation just as an exercise. Now I want to test if my feedforward implementation gives the expected output. I'm doing this manually right now and it's starting to get tedious, specially since the resulting topology is kind of crazy.

Is there some kind of dataset online which I can cross-reference? I want to do something like: given this input and topology+weights, this should be the output.

  • 1
    $\begingroup$ Can't you just build the same networks in pytorch/tensorflow, and compare the results? $\endgroup$
    – user114966
    Jan 14 at 18:34
  • $\begingroup$ Yeah I'm currently reading about these frameworks (haven't used any of them since I'm new to this area). Initial impression is that it seems rather overkill just to check the feedforward part. But yes, I could do that. Just need to familiarize myself first. $\endgroup$ Jan 14 at 18:45
  • $\begingroup$ I suspect using a standard framework will save you time over re-implementing everything from scratch -- it'll save you debugging the bugs in your own code, for instance. $\endgroup$
    – D.W.
    Jan 14 at 21:01
  • $\begingroup$ I understand. This is just a self-imposed exercise just for fun/learning. I do plan to eventually use these frameworks. $\endgroup$ Jan 14 at 21:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.