There is an official example, that we can set clusters number.
However, when I set the number to run the example, funny thing happens.
The same problem in ClusteringComponents
According to the theory of Spectral Clustering method, the feature number is the same to the eigenvectors sliced to use, which could be set by hand. Another reason is setting the number is the need for the final step of Spectral Clustering when using Kmeans method.
And this is also one feature of Spectral Clustering method.
So, what's the reasons for mathematica's strange behavior? Is it a bug? I'm on Mac version of 13.1.
Meanwhile, in GaussianMixture
method, we ca choose more components to clustering some data.
There is a same problem in ClusteringComponents
method.
FindClusters
introduced in Year 2007.
ClusteringComponents
introduced in Year 2010.
Now is Year 2022, I think many people may try this method.