Locally-scaled spectral clustering using empty region graphs.
Carlos D. CorreaPeter LindstromPublished in: KDD (2012)
Keyphrases
- spectral clustering
- graph partitioning
- graph construction
- graph clustering
- spectral methods
- pairwise
- laplacian matrix
- clustering method
- data clustering
- similarity graph
- graph laplacian
- image segmentation
- clustering algorithm
- k means
- constrained spectral clustering
- pairwise similarities
- eigendecomposition
- graph matching
- affinity matrix
- evolutionary clustering
- normalized cut
- similarity matrix
- adjacency matrix
- graph structures
- pairwise constraints
- graph kernels
- constrained clustering
- neighborhood graph
- machine learning
- graph mining
- weighted graph
- cluster analysis
- maximum likelihood
- feature selection