Partitioning clustering (k-means, k-kernel, etc.). Minimum k=2.

ShampunShampun Member Posts: 2 Contributor I
edited November 2018 in Help

Good day,

 

Why the for the partitioning clustering algorithms the minimum input parameter k=2 ? If i want to change it to 1, it's automatically set it to 2 (nota bene k-means and others are applicable for k=1...K). How then to know if the data set contains one cluster (without validation)?

 

Best regards,

TP

Answers

  • Thomas_OttThomas_Ott RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,761 Unicorn

    K=1 means you get one cluster, so that won't work. The default is K=2, for two clusters.

     

    If you want to find the optimal # of clusters, use the X-means operator. The defaults are a min of 2 and a max to 60.

Sign In or Register to comment.