Randomly Initialized Alternating Least Squares: Fast Convergence for Matrix Sensing.
Kiryung LeeDominik StögerPublished in: CoRR (2022)
Keyphrases
- alternating least squares
- negative matrix factorization
- nonnegative matrix factorization
- low rank approximation
- iterative algorithms
- stochastic gradient descent
- singular value decomposition
- matrix factorization
- least squares
- convergence rate
- spectral clustering
- low rank
- document clustering
- step size
- signal processing
- convergence speed
- sparse representation
- machine learning
- original data
- linear combination
- support vector machine
- feature space
- learning algorithm