Bi-stochastically normalized graph Laplacian: convergence to manifold Laplacian and robustness to outlier noise.
Xiuyuan ChengBoris LandaPublished in: CoRR (2022)
Keyphrases
- graph laplacian
- random walk
- spectral analysis
- heat kernel
- euclidean space
- basis functions
- spectral clustering
- laplacian eigenmaps
- manifold structure
- weighted graph
- geometrical structure
- pointwise
- kernel machines
- laplace beltrami
- underlying manifold
- neighborhood graph
- laplacian matrix
- linear combination
- semi supervised learning
- dimensionality reduction
- shape analysis
- low dimensional
- labeled and unlabeled data
- shape variations
- manifold learning
- missing data
- pairwise