Upper bound of Bayesian generalization error in non-negative matrix factorization.
Naoki HayashiSumio WatanabePublished in: Neurocomputing (2017)
Keyphrases
- generalization error
- negative matrix factorization
- upper bound
- lower bound
- matrix factorization
- classification error
- sample complexity
- document clustering
- training error
- sparse representation
- worst case
- sample size
- principal component analysis
- gradient method
- bayesian networks
- cross validation
- vc dimension
- model selection
- collaborative filtering
- spectral clustering
- learning algorithm
- neural network
- supervised learning
- feature vectors
- pattern recognition
- keywords