Vector Transport Free Riemannian LBFGS for Optimization on Symmetric Positive Definite Matrix Manifolds.
Reza GodazBenyamin GhojoghReshad HosseiniReza MonsefiFakhri KarrayMark CrowleyPublished in: ACML (2021)
Keyphrases
- symmetric positive definite
- riemannian manifolds
- vector space
- positive definite
- distribution function
- euclidean space
- lie group
- diffusion tensor
- geometric structure
- manifold learning
- mean shift
- von neumann
- feature space
- kernel matrix
- covariance matrix
- fisher information
- log euclidean
- human brain
- anisotropic diffusion
- metric space
- reproducing kernel hilbert space
- similarity search
- linear combination
- data points