Semidefinite Programming Based Preconditioning for More Robust Near-Separable Nonnegative Matrix Factorization.
Nicolas GillisStephen A. VavasisPublished in: CoRR (2013)
Keyphrases
- semidefinite programming
- nonnegative matrix factorization
- interior point
- primal dual
- linear programming
- matrix factorization
- interior point methods
- negative matrix factorization
- data representation
- linear program
- semidefinite
- least squares
- convergence rate
- kernel matrix
- collaborative filtering
- document clustering
- convex optimization
- approximation algorithms
- data mining
- original data
- image classification
- computer vision
- learning algorithm