Hi
I have a question regarding k-means clustering.What is the best way to find the value of K ?I have heard about silhouette coefficient but I am not sure how it works. I look forward to your help.
Thank you & kind regards,
Nandisha
There is no best way. The most popular ways are the elbow method and the silhouette method. Both allow you to evaluate a range of values of k.
A simple explanation can be found here.
The description of the Silhouette Coefficient node also is helpful.
1 Like
Hi @Nandisha,
There is also a Verified Componnet you can use here:
Optimized K-Means (Silhouette Coefficient)
Verified Components (knime.com/verified-components) are components made by the KNIME Team which work just like nodes! Drag and drop them from KNIME Hub to you KAP and add them to your workflow just like you would do with a standard node. If you want you can inspect the inside (right click Component > Open) or even customized them (Right Click > Component > Disconnect Link).
For more information about components:
Cheers
Paolo
1 Like
Hi Pablo,
thank you for the suggestion.Do you have an example of a workflow using this node so that I have an idea how it works?
Thank you & kind regards,
Nandisha
I just made one for you and I shared it on my space.
Enjoy!
2 Likes
That´s very nice of you.I appreciate.
Thank you so much.
Kind regards,
Nandisha
1 Like
This topic was automatically closed 182 days after the last reply. New replies are no longer allowed.