Spectral stability and perturbation results for kernel differentiation matrices on the sphere.
Thomas HangelbroekChristian RiegerGrady B. WrightPublished in: CoRR (2023)
Keyphrases
- positive definite
- positive semidefinite
- kernel function
- normalized cut
- kernel methods
- spectral methods
- kernel matrix
- eigendecomposition
- spectral features
- support vector
- laplacian matrix
- spectral analysis
- hyperspectral images
- input space
- pairwise comparison
- feature space
- kernel learning
- machine learning
- multiple kernel learning
- hyperspectral imagery
- similarity function
- unit sphere
- coefficient matrix
- pseudo inverse
- mean shift
- representer theorem