SNR threshold region prediction via singular value decomposition of the Barankin bound kernel.
John S. KotaAntonia Papandreou-SuppappolaPublished in: ACSSC (2017)
Keyphrases
- singular value decomposition
- singular vectors
- singular values
- least squares
- low rank
- signal to noise ratio
- latent semantic indexing
- principal component analysis
- kernel matrix
- low rank approximation
- dimension reduction
- kernel function
- kernel methods
- feature space
- data matrix
- dimensionality reduction
- positive definite
- pattern recognition
- qr decomposition
- data dependent
- input space
- input image