I have some input feature representing a physical quantity that's measured with a given incertitude.

Is there a way to feed the information about the incertitude to, let's say, a regressor using that feature in an effective way?

Two ways I see are creating new features by adding/subtracting the incertitude or making different predictions using the measured value and the upper/lower bounds separately. But I'm wondering if someone has a better approach up the sleeve!

  • $\begingroup$ You could try feeding quantiles, e.g. 5th, 25th, 50th, 75th, 95th $\endgroup$
    – Cryo
    Nov 12 at 12:36


Your Answer

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

Browse other questions tagged or ask your own question.