I am confused with the concepts of stratified sampling and oversampling for imbalanced datasets. From what I read in this question here: why-use-stratified-cross-validation-why-does-this-not-damage-variance-related , stratified sampling helps to make sure the sample contains the sample proportion of each class, so that the sample is representative.

But in many other posts I have read that oversampling rare events for imbalanced datasets is some helpful technique to handle high false positive or false negative problems.

Aren't the 2 sampling ideas contrary with each other? As in one is to make sure the sample is representative of the population, while the other is balancing the sample classes manually? I know I might confuse some fundamental concepts here. Can anyone help?


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.