Relative concentration bounds for the kernel matrix spectrum.
Ernesto Araya ValdiviaPublished in: CoRR (2018)
Keyphrases
- kernel matrix
- kernel methods
- semidefinite programming
- low rank
- metric learning
- kernel function
- kernel matrices
- feature space
- model selection
- kernel learning
- support vectors
- training samples
- input space
- upper bound
- ls svm
- reproducing kernel hilbert space
- neural network
- generalization bounds
- convex optimization
- matrix factorization
- distance metric
- dimensionality reduction
- multi class
- high dimensional
- feature extraction