Your data analysis solution
Name is required.
Email address is required.
Invalid email address
Answer is required.
Exceeding max length of 5KB

Agglomerative Hierarchical Clustering

By Simone Bregaglio | Jan 25, 2016 02:04PM CET

Dear developers,
I wonder if you could give an exhaustive description of the algorithm used to determine the number of clusters in the automatic truncation mode, when agglomerative hierarchical clustering is performed. I searched in many scientific papers in which XLSTAT is used, and the recurrent and standard sentence states that the "automatic truncation is based on the entropy and tries to create homogeneous groups". Could you give us further details?
Thank you in advance and best regards.




By Jean Paul | Jan 25, 2016 03:55PM CET | XLSTAT Agent

Dear XLSTAT user,

Thank you for contacting us.

The “entropy” option is based on the largest decrease in Shannon’s entropy between a node and the next one.
XLSTAT minimizes 1/(entropy(node i-1) -entropy(node(i))). When the minimum is reached, the truncation is applied.

The “inertia” option is based on the largest difference in inertia (simply measured on the dendrogram’s Y axis) between (node i) and (node i-1).

I hope this helps!


Jean Paul

This question has received the maximum number of answers.

Contact Us
seconds ago
a minute ago
minutes ago
an hour ago
hours ago
a day ago
days ago
Invalid characters found