RaCT: Toward Amortized Ranking-Critical Training For Collaborative Filtering.
Sam LobelChunyuan LiJianfeng GaoLawrence CarinPublished in: ICLR (2020)
Keyphrases
- collaborative filtering
- recommender systems
- maximum margin matrix factorization
- learning to rank
- training set
- matrix factorization
- web search
- ranking functions
- training phase
- test set
- training process
- training examples
- ranking algorithm
- neural network
- supervised learning
- link analysis
- recommendation systems
- probabilistic matrix factorization
- running times
- search tree
- learning algorithm
- data sparsity
- personalized recommendation
- worst case