I want to train a multi-class classification deep learning model. But my dataset is class-imbalanced. So considering 2 solutions, random oversampling and classes weighting, I have some questions:

  1. Do random oversampling and classes weighting have the same effect on the model training? If they don't, what are the differences?
  2. Which method is better? Why?
  • 1
    $\begingroup$ Better for what? What problem are you trying to solve? $\endgroup$ – Sycorax Jun 15 '19 at 13:17
  • $\begingroup$ I edited the question. $\endgroup$ – Kyouko Jun 17 '19 at 6:45

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.