Matrix completion and low-rank SVD via fast alternating least squares.
Trevor HastieRahul MazumderJason D. LeeReza ZadehPublished in: J. Mach. Learn. Res. (2015)
Keyphrases
- low rank
- matrix completion
- low rank approximation
- singular value decomposition
- convex optimization
- low rank matrix
- singular values
- kernel matrix
- missing data
- linear combination
- matrix factorization
- rank minimization
- high dimensional data
- semi supervised
- reconstruction error
- least squares
- eigendecomposition
- high order
- principal component analysis
- latent semantic indexing
- stochastic gradient descent
- trace norm
- collaborative filtering
- negative matrix factorization
- small number
- training set
- support vector
- training data