Matrix Sparsification and the Sparse Null Space Problem.

Lee-Ad Gottlieb Tyler Neylon

Published in: APPROX-RANDOM (2010)

Keyphrases

null space
theoretical guarantees
linear discriminant analysis
projection matrix
singular value decomposition
novelty detection
principal components

singular values
sparse representation
high dimensional
scatter matrices
least squares
discriminant analysis
discriminative information

learning algorithm
small sample size
random projections
dimension reduction
image processing
finite dimensional

natural images
dimensionality reduction
intrinsic dimensionality
worst case
object recognition
data sets