idx = spectralcluster(X, k)
idx = spectralcluster(S, k, 'Distance', 'precomputed')
idx = spectralcluster(..., Name, Value)
[idx, V, D] = spectralcluster(...)
spectralcluster partitions observations by constructing a similarity graph, computing a graph Laplacian embedding, and clustering the embedded rows.
Name-value arguments include Distance, P, Cov, Scale, SimilarityGraph, NumNeighbors, KNNGraphType, Radius, KernelScale, LaplacianNormalization, and ClusterMethod.
X = [0 0; 0 1; 1 0; 10 10; 10 11; 11 10];
idx = spectralcluster(X, 2, 'NumNeighbors', 5, 'KernelScale', 2)| Version | Description |
|---|---|
| 2.0.0 | initial version |