Tight convex relaxations for sparse matrix factorization.
Emile RichardGuillaume ObozinskiJean-Philippe VertPublished in: CoRR (2014)
Keyphrases
- matrix factorization
- convex relaxation
- tensor factorization
- collaborative filtering
- recommender systems
- latent factors
- convex optimization
- low rank
- nonnegative matrix factorization
- negative matrix factorization
- globally optimal
- missing data
- factorization methods
- multi label
- multistage
- optimization methods
- multiple kernel learning
- image classification
- object recognition
- feature selection
- binary matrix
- sparsity constraints
- sparse coding
- sparse representation
- nearest neighbor
- feature vectors
- feature space
- face recognition