Graph-Laplacian PCA: Closed-Form Solution and Robustness.
Bo JiangChris H. Q. DingBin LuoJin TangPublished in: CVPR (2013)
Keyphrases
- closed form solutions
- graph laplacian
- principal component analysis
- random walk
- spectral clustering
- point correspondences
- spectral analysis
- closed form
- euclidean space
- basis functions
- weighted graph
- kernel pca
- pointwise
- kernel machines
- feature extraction
- constrained optimization
- low dimensional
- manifold structure
- neighborhood graph
- special case
- high dimensional
- feature space
- machine learning
- face recognition
- random projections
- semi supervised learning
- feature points
- covariance matrix
- clustering method
- labeled and unlabeled data
- markov chain
- dimensionality reduction
- semi supervised
- viewpoint