k-means++: few more steps yield constant approximation.
Davin ChooChristoph GrunauJulian PortmannVáclav RozhonPublished in: CoRR (2020)
Keyphrases
- k means
- clustering algorithm
- approximation error
- approximation algorithms
- data clustering
- spectral clustering
- closed form
- self organizing maps
- cluster analysis
- expectation maximization
- error bounds
- approximation methods
- constant factor
- database
- hierarchical clustering
- cluster centers
- approximation ratio
- sufficient statistics
- fuzzy c means
- search algorithm
- multiscale
- exact computation
- axis parallel