How to use K-Centroids Diagnostics output w/ K-Centroid Cluster Analysis Tool

[CoreyS]

Hello my fellow Alteryx Community members,

I need your help! 

I'm attempting to use the K-Centroids Diagnostics output results w/ K-Centroid Cluster Analysis Tool. However, I don't know how to interpret the Diagnostic results (i.e. K-Means Cluster Assessment Report) and use them to perform the Cluster Analysis.

What I'm confused about...

• What should I do with the Diagnostic output results (i.e. K-Means Cluster Assessment Report)?
• Are K-Centroid Diagnostics recommended as best practice before using K-Centroids Cluster Analysis Tool?

By the way, I did read the Alteryx published 'K-Centroids Diagnostics Tool' overview but still trying to understand its use at the basic level.

Below are the Diagnostics output results (i.e. K-Means Cluster Assessment Report):

Thanks for your help in advance!

Respectfully,

Corey

[ianwi · Accepted solution]

Hi @CoreyS,

I would suggest also reviewing the Predictive Grouping interactive lessons in the Data Science Learning Path, specifically Clustering in Designer.

To answer your question, the K-Centroids Diagnostics tool can be used to determine the most appropriate number of clusters for your dataset. The interpretation of the report can be a little subjective but it helps you to narrow down the choices before configuring the number of clusters in the K-Centroids Cluster Analysis tool.

Hope this helps!

[CoreyS]

@ianwi  thanks for the informative references. I'll definitely check out the "Data Science Learning Path/Clustering in Designer".

Respectfully,

Corey