Distribution-free uncertainty quantification for kernel methods by gradient perturbations.
Balázs Csanád CsájiKrisztián Balázs KisPublished in: Mach. Learn. (2019)
Keyphrases
- kernel methods
- distribution free
- learning problems
- normal distribution
- kernel function
- sample complexity
- machine learning
- large deviations
- support vector
- vc dimension
- support vector machine
- feature space
- kernel matrix
- learning tasks
- reproducing kernel hilbert space
- active learning
- concept class
- sample size
- kernel machines