I want to know if there is a measure of how well two classes in Y are separable (linearly or not) based on their features in X. Easiest way of explaining this is to compare it to correlation coefficients, the higher the correlation the higher possiblity for successful regression based on given feature (at least in theory).

Is there any measure that will tell me how well classes are separated based on input data features, before training a ML model?

  • $\begingroup$ You can use a different measure for the decision tree. Anyhow, it's an ML model. $\endgroup$
    – OmG
    Oct 20 '19 at 21:18

Are you thinking something like Information Gain?

Information Gain basically uses the concept of information entropy to determine if splitting a variable is useful.


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.