Positive semi-definite embedding for dimensionality reduction and out-of-sample extensions.
Michaël FanuelAntoine AspeelJean-Charles DelvenneJohan A. K. SuykensPublished in: CoRR (2017)
Keyphrases
- dimensionality reduction
- positive semi definite
- metric learning
- nonlinear dimensionality reduction
- graph embedding
- graph kernels
- low dimensional
- data representation
- principal component analysis
- high dimensional
- manifold learning
- kernel methods
- kernel function
- high dimensional data
- feature extraction
- pattern recognition
- feature space
- vector space
- similarity function
- feature selection
- data points
- kernel pca
- dot product
- input space
- singular value decomposition
- random projections
- kernel matrix
- linear discriminant analysis
- label information
- pairwise
- distance metric
- learning tasks
- euclidean distance