Sparse low rank factorization for deep neural network compression.
Sridhar SwaminathanDeepak GargRajkumar KannanFrédéric AndrèsPublished in: Neurocomputing (2020)
Keyphrases
- low rank
- neural network
- low rank matrix
- low rank subspace
- low rank matrices
- rank minimization
- nuclear norm
- missing entries
- sparsity constraints
- matrix factorization
- low rank representation
- robust principal component analysis
- convex optimization
- singular value decomposition
- missing data
- linear combination
- regularized regression
- matrix completion
- kernel matrices
- low rank approximation
- matrix decomposition
- semi supervised
- tensor decomposition
- high dimensional data
- high order
- kernel matrix
- data matrix
- non rigid structure from motion
- image compression
- minimization problems
- pattern recognition
- factorization methods
- sparse representation
- dimensionality reduction
- affinity matrix
- singular values
- low rank and sparse
- high dimensional
- data sets