NysADMM: faster composite convex optimization via low-rank approximation.
Shipu ZhaoZachary FrangellaMadeleine UdellPublished in: ICML (2022)
Keyphrases
- convex optimization
- low rank approximation
- low rank
- kernel matrix
- low rank matrix
- interior point methods
- matrix completion
- singular value decomposition
- total variation
- convex relaxation
- subspace learning
- norm minimization
- nonnegative matrix factorization
- machine learning
- iterative algorithms
- eigendecomposition
- singular values
- partial differential equations
- adjacency matrix
- image restoration
- least squares
- computer vision