The use of k-means++ for approximate spectral clustering of large datasets.
Berna YalçinKadim TasdemirPublished in: SIU (2014)
Keyphrases
- spectral clustering
- k means
- clustering method
- data clustering
- pairwise
- clustering algorithm
- evolutionary clustering
- clustering approaches
- graph laplacian
- similarity matrix
- clustering quality
- normalized cut
- image segmentation
- eigendecomposition
- graph partitioning
- affinity matrix
- nonnegative matrix factorization
- pairwise constraints
- cluster analysis
- constrained clustering
- data sets
- unsupervised clustering
- expectation maximization
- laplacian matrix
- similarity graph
- alternative clusterings
- constrained spectral clustering