Noise Regularizes Over-parameterized Rank One Matrix Recovery, Provably.
Tianyi LiuYan LiEnlu ZhouTuo ZhaoPublished in: AISTATS (2022)
Keyphrases
- random noise
- singular values
- nuclear norm minimization
- noise level
- singular value decomposition
- noisy data
- gaussian noise
- image recovery
- noise model
- additive noise
- convex optimization
- noise reduction
- signal to noise ratio
- low rank
- learning algorithm
- ranked list
- image structure
- image noise
- linear algebra
- low signal to noise ratio
- noise sensitivity
- lower bound