Always run K-Finder before using K-Means, Gaussian Mixture, or Spectral Clustering. The suggested K value is a recommendation you may want to try nearby values (K±1) and compare results using the Cluster Quality node.
Shapely Guide
Geometry
Machine Learning Guide
Analysis
K-Finder
K-Finder automatically determines the optimal number of clusters (K) for your data using methods like the Elbow Method and Silhouette Analysis. It tests different K values and recommends the best one.
This removes the guesswork from choosing K for algorithms like K-Means and Gaussian Mixture.
⌘I