I want to get the Principal components of a dataset and apply K mean clustering on them. Do I need to Normalized the PCA output before applying Kmeans on them ?

  • $\begingroup$ If the ranges of the numeric attributes differ by many orders of magnitude, then yes. You could also scale (divide each by standard deviation in that column for instance) $\endgroup$
    – knb
    Apr 11 '18 at 10:56
  • $\begingroup$ Also you must scale the data before applying PCA. I hope you know that! $\endgroup$
    – spectre
    2 days ago

No - There is no need to normalize after Principal Component Analysis (PCA) because each dimension is on the same scale.


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.