Simultaneous Discovery of Common and Discriminative Topics via Joint Nonnegative Matrix Factorization.
Hannah KimJaegul ChooJingu KimChandan K. ReddyHaesun ParkPublished in: KDD (2015)
Keyphrases
- nonnegative matrix factorization
- probabilistic latent semantic indexing
- negative matrix factorization
- latent dirichlet allocation
- matrix factorization
- data representation
- least squares
- topic models
- spectral clustering
- principal component analysis
- nonnegative matrix
- semi supervised
- information retrieval
- document clustering
- knowledge discovery
- original data
- object recognition
- feature extraction
- feature selection
- sparse representation
- missing data
- unsupervised learning
- information extraction
- image retrieval
- pattern recognition
- discriminative learning
- sparsity constraints
- data mining