Trading performance for memory in sparse direct solvers using low-rank compression.
Loris MarchalThibault MaretteGrégoire PichonFrédéric VivienPublished in: Future Gener. Comput. Syst. (2022)
Keyphrases
- low rank
- low rank matrix
- rank minimization
- low rank matrices
- low rank subspace
- nuclear norm
- sparsity constraints
- low rank representation
- robust principal component analysis
- sparse linear
- linear combination
- group sparsity
- matrix factorization
- regularized regression
- matrix completion
- kernel matrices
- convex optimization
- missing data
- low rank approximation
- high dimensional data
- singular value decomposition
- matrix decomposition
- kernel matrix
- high order
- semi supervised
- singular values
- tensor decomposition
- image compression
- sparse matrix
- high dimensional
- minimization problems
- affinity matrix
- trace norm
- data matrix
- collaborative filtering
- binary matrices
- structured sparsity
- data sets
- spectral clustering
- pattern recognition