Connecting the Dots - Density-Connectivity Distance unifies DBSCAN, k-Center and Spectral Clustering.
Anna BeerAndrew DraganovEllen HohmaPhilipp JahnChristian M. M. FreyIra AssentPublished in: KDD (2023)
Keyphrases
- spectral clustering
- data clustering
- clustering method
- k means
- clustering algorithm
- instance level constraints
- density based clustering algorithm
- pairwise
- graph laplacian
- similarity matrix
- affinity matrix
- evolutionary clustering
- dense regions
- cluster analysis
- clustering quality
- normalized cut
- eigendecomposition
- euclidean distance
- spectral methods
- constrained spectral clustering
- image segmentation
- similarity graph
- subspace clustering
- constrained clustering
- label propagation
- graph partitioning
- machine learning
- document clustering
- unsupervised learning
- distance measure
- similarity measure
- image processing
- computer vision