Interpretable Discriminative Dimensionality Reduction and Feature Selection on the Manifold.
Babak HosseiniBarbara HammerPublished in: CoRR (2019)
Keyphrases
- dimensionality reduction
- feature selection
- manifold learning
- low dimensional
- lower dimensional
- feature extraction
- feature space
- diffusion maps
- nonlinear dimensionality reduction
- class discrimination
- graph embedding
- locally linear embedding
- unsupervised learning
- manifold structure
- class separability
- high dimensionality
- nonlinear manifold
- high dimensional
- latent space
- discriminative features
- high dimensional data
- locality preserving
- subspace learning
- data representation
- linear discriminant analysis
- dimension reduction
- dimensionality reduction methods
- text categorization
- principal component analysis
- classification accuracy
- support vector machine
- feature set
- multi class
- pattern recognition
- feature selection algorithms
- random projections
- euclidean distance
- locality preserving projections
- high dimension
- metric learning
- underlying manifold
- discriminant information
- unsupervised feature selection
- informative features
- embedding space
- principal components
- euclidean space
- model selection
- data points
- multidimensional scaling
- linear subspace
- sparse representation
- geodesic distance
- machine learning
- intrinsic dimensionality
- kernel learning
- discriminative learning
- feature subset
- pairwise
- linear dimensionality reduction
- multi task