Improving the recovery of principal components with semi-deterministic random projections.
Keegan KangGiles HookerPublished in: CISS (2016)
Keyphrases
- principal components
- random projections
- dimensionality reduction
- principal component analysis
- dimension reduction
- compressive sensing
- low dimensional
- high dimensional
- face recognition
- linear discriminant analysis
- covariance matrix
- high dimensional data
- feature space
- original data
- feature extraction
- unsupervised learning
- pattern recognition
- sparse representation
- feature selection
- high dimensionality
- singular value decomposition
- image reconstruction
- random sampling
- kernel principal component analysis
- data points
- similarity search
- hash functions
- principal component regression
- document clustering
- feature set
- preprocessing
- objective function
- multiscale
- image processing