Small width, low distortions: quasi-isometric embeddings with quantized sub-Gaussian random projections.
Laurent JacquesPublished in: CoRR (2015)
Keyphrases
- random projections
- low dimensional
- dimensionality reduction
- manifold learning
- compressive sensing
- dimension reduction
- euclidean space
- compressed sensing
- high dimensional
- high dimensional data
- image reconstruction
- original data
- data points
- random sampling
- principal component analysis
- text classification
- sparse representation
- feature space
- neural network
- high dimensionality
- digital libraries
- pattern recognition
- training data
- feature extraction