I have dataset with two label class (good and bad), I want to apply K Means on my dataset using python, should I use that label dataset or I have to delete the label class column ?

  • 1
    $\begingroup$ K-means clustering is done to give labels to data. You already have those, so why are you applying k-means? What is the problem statement? $\endgroup$ – bkshi Feb 9 '19 at 8:58
  • $\begingroup$ I think the OP actually meant the dataset contains a binary feature. $\endgroup$ – Louis T Feb 9 '19 at 9:49
  • $\begingroup$ Possible duplicate of K-Means clustering for mixed numeric and categorical data $\endgroup$ – Louis T Feb 9 '19 at 9:50
  • $\begingroup$ Thanks all for your answers , yes my dataset consists of binary features and I want to use clustering to test new samples $\endgroup$ – lona Feb 11 '19 at 3:23

Delete the label column.

Assuming that you want to compare the clusters to the labels later, then the labels must not be part of the data passed to k-means.

And k-means only works well on continuous variables anyway.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.