Threshold setting in Dataiku for Hierarchical Clustering

Punit_Shah
Punit_Shah Registered Posts: 1 ✭✭✭

Hi,

In hierarchical clustering one of the key concept is threshold of distance (euclidian) which is also key differentiator from K-Means clustering. This threshold value determines which clusters to merge and which not.

In DSS, I am unable to find this setting and it's just asking me to provide number of clusters. So wanted to check how it's different in K-Means in DSS and in which ways I can provide this threshold to my algorithm.

Thank you.

Tagged:

Answers

  • Ouma
    Ouma Dataiku DSS Core Designer, Dataiku DSS ML Practitioner, Dataiku DSS Core Concepts, Registered Posts: 12 ✭✭✭✭

    Hey,

    I'm actually having the same question, and I would like to know also which method of linkage is used for the Agglomerative clustering algo ?

    Thanks

Setup Info
    Tags
      Help me…