Modularity-based model selection for kernel spectral clustering.
Rocco LangoneCarlos AlzateJohan A. K. SuykensPublished in: IJCNN (2011)
Keyphrases
- model selection
- spectral clustering
- normalized cut
- kernel matrix
- cross validation
- hyperparameters
- data clustering
- pairwise
- clustering method
- motion segmentation
- graph laplacian
- sample size
- k means
- kernel machines
- clustering algorithm
- regression model
- similarity matrix
- gaussian process
- gaussian processes
- mixture model
- image segmentation
- machine learning
- kernel function
- generalization error
- kernel methods
- variable selection
- learning machines
- selection criterion
- pairwise constraints
- kernel learning
- model selection criteria
- similarity measure
- feature selection
- reproducing kernel hilbert space
- graph cuts
- text classification