Convergence to Second-Order Stationarity for Non-negative Matrix Factorization: Provably and Concurrently.
Ioannis PanageasStratis SkoulakisAntonios VarvitsiotisXiao WangPublished in: CoRR (2020)
Keyphrases
- negative matrix factorization
- update rule
- nonnegative matrix factorization
- matrix factorization
- sparse representation
- principal component analysis
- document clustering
- non stationary
- convergence rate
- spectral clustering
- tensor factorization
- constrained least squares
- machine learning
- clustering algorithm
- recommender systems
- learning rate