Online algorithms for nonnegative matrix factorization with the Itakura-Saito divergence.
Augustin LefèvreFrancis R. BachCédric FévottePublished in: WASPAA (2011)
Keyphrases
- nonnegative matrix factorization
- online algorithms
- bregman divergences
- relative entropy
- online learning
- learning algorithm
- negative matrix factorization
- matrix factorization
- lower bound
- data representation
- worst case
- least squares
- average case
- principal component analysis
- spectral clustering
- document clustering
- original data
- data mining
- cost sensitive
- linear space
- active learning
- pattern recognition
- data sets
- asymptotically optimal