Improved SVD-based initialization for nonnegative matrix factorization using low-rank correction.
Syed Muhammad AtifSameer QaziNicolas GillisPublished in: Pattern Recognit. Lett. (2019)
Keyphrases
- nonnegative matrix factorization
- low rank
- matrix decomposition
- matrix factorization
- singular value decomposition
- data matrix
- negative matrix factorization
- least squares
- sparsity constraints
- missing data
- collaborative filtering
- principal component analysis
- low rank matrix
- factorization methods
- recommender systems
- semi supervised
- matrix completion
- singular values
- convex optimization
- data representation
- interior point
- low rank approximation
- high order
- high dimensional data
- unsupervised learning
- k means
- image processing
- latent structure
- signal processing
- dimensionality reduction
- objective function