Projected Wirtinger Gradient Descent for Low-Rank Hankel Matrix Completion in Spectral Compressed Sensing.
Jian-Feng CaiSuhui LiuWeiyu XuPublished in: CoRR (2015)
Keyphrases
- compressed sensing
- matrix completion
- low rank
- image reconstruction
- missing data
- convex optimization
- linear combination
- random projections
- low rank matrix
- matrix factorization
- singular value decomposition
- sparse representation
- rank minimization
- high dimensional data
- natural images
- trace norm
- semi supervised
- high order
- stochastic gradient descent
- signal processing
- singular values
- kernel matrix
- collaborative filtering
- incomplete data
- loss function
- pattern recognition
- small number
- dimensionality reduction
- total variation
- rank aggregation
- dimension reduction
- multiscale
- super resolution
- feature selection