Hi. My dependent variable is imbalanced so I am doing an oversampling of the minority class. I will train my model on an oversampled training set. Do I test the model on the imbaanced data or the balanced data? Why
I say both.
because if you dont look at the imbalanced data, then you cant say anything about your models performance.
if you dont look at the balanced data you won't as easily diagnose problems if they crop up in the imbalanced data.
This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.
If you have a query related to it or one of the replies, start a new topic and refer back with a link.