Is the 1-norm the best convex sparse regularization?
Yann TraonmilinSamuel VaiterRémi GribonvalPublished in: CoRR (2018)
Keyphrases
- mixed norm
- sparsity inducing
- nuclear norm
- norm minimization
- structured sparsity
- convex optimization
- group lasso
- rank minimization
- norm regularization
- variable selection
- low rank
- minimization problems
- low rank matrices
- efficient optimization
- trace norm
- sparsity regularization
- sparse learning
- sparse regression
- matrix completion
- convex optimization problems
- total least squares
- sparse representation
- regression model
- regularization method
- convex functions
- regularized least squares
- robust principal component analysis
- compressive sensing
- penalty term
- regularization methods
- elastic net
- canonical correlation analysis
- alternating direction
- convex relaxation
- quasiconvex
- convex hull
- augmented lagrangian
- linear program
- sparse coding
- high dimensional
- total variation
- matrix factorization
- piecewise linear
- regularization term
- kernel matrix
- kernel matrices
- low rank matrix
- inverse problems
- low rank approximation