The target variable for my regression problem has a very high noise, so measurement error is very high and trends can only be seen in longer time periods.

What are good approaches to address this problem. Is it a good idea to directly try to predict the noisy target and hope that the model don't learn the noise? Or should the target variable be preprocessed in advance using filtering methods / moving averages, defining bins / classes or any other approaches?


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.