Useful statistical knowledge for clustering

Useful statistical knowledge for clustering

I wish I had learned earlier statistical techniques to assess characteristics and quality of a machine learning clustering technique i.e.:

– discriminant analysis,

– entropy to evaluate the spread of a variable within a cluster,

– Silhouette score and Silhouette shape.

By: Valérie Alléger

Leave a Reply

Your email address will not be published. Required fields are marked *