Uniform Deviation Bounds for Unbounded Loss Functions like k-Means.
Olivier BachemMario LucicS. Hamed HassaniAndreas KrausePublished in: CoRR (2017)
Keyphrases
- loss function
- k means
- pairwise
- upper bound
- support vector
- loss minimization
- squared error
- empirical risk
- risk minimization
- square loss
- lower bound
- clustering algorithm
- learning to rank
- reproducing kernel hilbert space
- data clustering
- loss bounds
- convex loss functions
- spectral clustering
- hinge loss
- uniform convergence
- clustering method
- similarity measure
- worst case