Fast Parallel Randomized Algorithm for Nonnegative Matrix Factorization with KL Divergence for Large Sparse Datasets.
Duy Khuong NguyenTu Bao HoPublished in: CoRR (2016)
Keyphrases
- nonnegative matrix factorization
- kl divergence
- randomized algorithm
- negative matrix factorization
- kullback leibler divergence
- matrix factorization
- mahalanobis distance
- data representation
- least squares
- lower bound
- spectral clustering
- probabilistic latent semantic analysis
- high dimensional
- sparse representation
- exponential family
- maximum likelihood
- principal component analysis
- collaborative filtering
- objective function
- original data
- document clustering
- probabilistic model