Stability of convergence rates: Kernel interpolation on non-Lipschitz domains.
Tizian WenzelGabriele SantinBernard HaasdonkPublished in: CoRR (2022)
Keyphrases
- convergence rate
- gaussian kernels
- numerical stability
- learning rate
- convergence speed
- kernel function
- global convergence
- reproducing kernel hilbert space
- conjugate gradient
- learning algorithm
- kernel methods
- feature space
- support vector
- machine learning
- kernel matrix
- kernel learning
- kernel machines
- hilbert space
- primal dual
- gaussian kernel
- stopping criterion