A probabilistic model selection criterion for spectral clustering.
Pierrick BruneauBenoît OtjacquesPublished in: Intell. Data Anal. (2018)
Keyphrases
- spectral clustering
- pairwise
- clustering method
- data clustering
- similarity matrix
- clustering algorithm
- image segmentation
- affinity matrix
- eigendecomposition
- graph partitioning
- nonnegative matrix factorization
- evolutionary clustering
- probabilistic model
- constrained clustering
- constrained spectral clustering
- generative model
- spectral methods
- bayesian networks
- negative matrix factorization
- clustering quality
- normalized cut
- pairwise constraints
- alternative clusterings
- uncertain data
- inference process
- reinforcement learning
- laplacian matrix