Let's say I have a data set with of length N. A small proportion N2 is labeled. Can I remove some labels and then 'reverse' this action with a trained neural network? I could then use the same process to fill the other (N - N2) rows with labels.

  • 1
    $\begingroup$ What is the purpose of your trained neural network? Is it to label the data like N2 is labelled? Or is it to provide new labels? $\endgroup$ Feb 20 '20 at 9:13
  • $\begingroup$ purpose is to label to have N labels. $\endgroup$ Feb 21 '20 at 10:10
  • $\begingroup$ N different labels? So each data point has a different label? Or do you mean each data point is labeled with a label already existing in N2? $\endgroup$ Feb 21 '20 at 10:13
  • $\begingroup$ not N labels, let's say 2 labels. N2 rows have 2 labels, the other N - N2 have no labels. $\endgroup$ Feb 24 '20 at 14:10
  • $\begingroup$ How is this different from supervised learning, trained on your labeled data, and used on your unlabeled data? $\endgroup$ Feb 26 '20 at 7:55

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.