Alternating Linear Bandits for Online Matrix-Factorization Recommendation.
Hamid DadkhahiSahand NegahbanPublished in: CoRR (2018)
Keyphrases
- matrix factorization
- collaborative filtering
- recommender systems
- item recommendation
- data sparsity
- latent factor models
- cold start problem
- low rank
- negative matrix factorization
- online learning
- low rank matrix
- factor analysis
- nonnegative matrix factorization
- implicit feedback
- personalized recommendation
- variational bayesian
- factorization methods
- rating prediction
- missing data
- data matrix
- transfer learning
- probabilistic matrix factorization
- user ratings
- stochastic gradient descent
- tensor factorization
- multiscale
- recommendation systems
- closed form
- binary matrix
- user preferences