Two-stage convex relaxation approach to least squares loss constrained low-rank plus sparsity optimization problems.
Le HanShujun BiShaohua PanPublished in: Comput. Optim. Appl. (2016)
Keyphrases
- low rank
- convex relaxation
- convex optimization
- least squares
- optimization problems
- sparse approximation
- matrix completion
- singular value decomposition
- optimization methods
- linear combination
- evolutionary algorithm
- cost function
- low rank matrix
- interior point methods
- total variation
- nuclear norm
- kernel matrix
- norm minimization
- matrix factorization
- optical flow
- multi label
- trace norm
- objective function
- sparse representation
- missing data
- globally optimal
- multiple kernel learning
- multistage
- high order
- small number
- ls svm
- high dimensional
- pairwise
- face recognition
- feature selection
- learning algorithm