Pairwise Preferences Based Matrix Factorization and Nearest Neighbor Recommendation Techniques.
Saikishore KallooriFrancesco RicciMarko TkalcicPublished in: RecSys (2016)
Keyphrases
- matrix factorization
- nearest neighbor
- pairwise
- collaborative filtering
- recommender systems
- user preferences
- item recommendation
- data sparsity
- cold start problem
- low rank
- latent factor models
- knn
- recommendation systems
- nonnegative matrix factorization
- training set
- high dimensional data
- rating prediction
- negative matrix factorization
- markov random field
- user behavior
- personalized recommendation
- higher order
- factorization methods
- implicit feedback
- personalized ranking
- probabilistic matrix factorization
- factor analysis
- loss function
- high dimensional
- high order
- missing data
- tensor factorization
- similarity measure
- recommendation algorithms
- user profiles
- variational bayesian
- user ratings
- stochastic gradient descent
- latent factors
- semi supervised
- feature extraction
- face recognition