Wasserstein Gradient Flows for Moreau Envelopes of f-Divergences in Reproducing Kernel Hilbert Spaces.
Sebastian NeumayerViktor SteinGabriele SteidlPublished in: CoRR (2024)
Keyphrases
- reproducing kernel hilbert space
- kernel methods
- loss function
- learning theory
- real valued
- active contour model
- density estimation
- kernel function
- euclidean space
- special case
- domain adaptation
- learning problems
- distance measure
- data dependent
- gaussian process
- input space
- finite dimensional
- riemannian manifolds
- kernel matrix
- metric learning
- energy functional
- linear model
- region growing
- active contours