Share your ideas for the JMP Scripting Unsession at Discovery Summit by September 17th. We hope to see you there!
Choose Language Hide Translation Bar
0 Kudos

Nonhierarchical K-means Clustering

Can we have JMP clustering platform use the Ward's hierarchical method as starting solution for k-means clustering? Since the objective of the two methods are similar (minimizing the within cluster variance) this strategy should prove useful.  There is external research that supports this (Steinley and Brusco (2007)).   According to their research, the only other way to improve is to increase the number of random initializations to 10000 and beyond.  If that is so, can JMP give the option for users to control the number of iterations?

Tracking Number:

Defect ID: