Pretty simple question here:

Is it useful to use the standard deviation, skew, kurtosis, or any other extrapolatory stats as features, and if so in which problem sets?

In this case, I am talking about deep learning problems.


I would say it is useful if you have an extensive knowledge in the domain you want to apply your model in. You also need more data for it to yield reasonable results.
As for real world uses l can only think of trading at the moment.

| improve this answer | |
  • $\begingroup$ Can you give me an example of where in trading it'd be used. I.e. as an observation in an RL agent used for crypto trading. $\endgroup$ – hisairnessag3 May 17 '18 at 21:47

Your Answer

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

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