Low-rank regularization and solution uniqueness in over-parameterized matrix sensing.
Kelly GeyerAnastasios KyrillidisAmir KalevPublished in: AISTATS (2020)
Keyphrases
- low rank
- trace norm
- rank minimization
- convex optimization
- norm minimization
- singular value decomposition
- matrix completion
- linear combination
- low rank matrix
- missing data
- matrix decomposition
- matrix factorization
- group sparsity
- minimization problems
- kernel matrices
- nuclear norm
- high order
- low rank approximation
- semi supervised
- high dimensional data
- singular values
- kernel matrix
- frobenius norm
- factorization methods
- multi task
- data matrix
- least squares
- robust principal component analysis
- learning algorithm
- low rank and sparse
- affinity matrix
- higher order
- tensor decomposition
- approximation methods
- machine learning