A fast non-convex regularizer for low rank matrix completion.
Cho-Ying WuJian-Jiun DingPublished in: APSIPA (2017)
Keyphrases
- matrix completion
- low rank
- convex optimization
- nuclear norm
- total variation
- semi supervised
- convex relaxation
- minimization problems
- low rank matrix
- rank minimization
- low rank matrices
- matrix factorization
- linear combination
- singular value decomposition
- missing data
- singular values
- interior point methods
- image denoising
- norm minimization
- regularization term
- kernel matrix
- image restoration
- denoising
- collaborative filtering
- high dimensional data
- affinity matrix
- high order
- globally optimal
- incomplete data
- higher order
- learning algorithm
- trace norm
- multi label
- training set
- feature extraction