A New Approach of Matrix Factorization and Its Application in Recommender Systems.
Hong PengShuyi HongLinkai LuoQifeng ZhouXiaoqin HuangPublished in: ICMLA (2016)
Keyphrases
- matrix factorization
- recommender systems
- collaborative filtering
- low rank
- data sparsity
- factorization methods
- cold start problem
- nonnegative matrix factorization
- item recommendation
- negative matrix factorization
- user preferences
- probabilistic matrix factorization
- cold start
- data matrix
- implicit feedback
- factor analysis
- variational bayesian
- recommendation quality
- user feedback
- information overload
- recommendation systems
- user interests
- personalized ranking
- recommendation algorithms
- factorization method
- binary matrix
- user behavior
- natural images