2

I made K-means clustering to a large dataset with 32000 features,

Ps : I cant delete outliers because I want to detect them and make a cluster that contains anomalies in my data ...

--> My Question is :

  • Is it normal for the ditribution to be like this?
  • Is there any other way to analyse kmeans clusters ?

pairplot graph

data_ba
  • 31
  • 3
  • 1
    I don't think you are supposed to use pair-plot on the clustered results. Do the pair-plot before, as a EDA exercise. Then, do your clustering afterwards. See the link below for one good example of how this could be done. https://statisticallyrelevant.com/k-means-clustering-in-python/ – ASH Jul 11 '22 at 13:42

0 Answers0