Bayesian Cramer-Rao bounds for factorized model based low rank matrix reconstruction.
Martin SundinSaikat ChatterjeeMagnus JanssonPublished in: EUSIPCO (2016)
Keyphrases
- low rank matrix
- matrix factorization
- low rank
- approximation methods
- lower bound
- upper bound
- convex optimization
- singular value decomposition
- collaborative filtering
- sparse matrix
- image reconstruction
- recommender systems
- worst case
- high resolution
- bayesian networks
- fisher information
- linear combination
- missing data
- maximum likelihood
- error bounds
- high dimensional data
- machine learning
- semi supervised
- active learning
- high dimensional
- parameter estimates
- pattern recognition
- face recognition
- computer vision