Split gradient method for nonnegative matrix factorization.
Henri LantériCéline TheysCédric RichardCédric FévottePublished in: EUSIPCO (2010)
Keyphrases
- gradient method
- nonnegative matrix factorization
- negative matrix factorization
- matrix factorization
- document clustering
- sparse representation
- principal component analysis
- convergence rate
- data representation
- data sets
- spectral clustering
- least squares
- text mining
- optimization methods
- step size
- dimensionality reduction
- low dimensional
- image representation
- kernel function
- pairwise
- objective function
- machine learning