A PAC bound for joint matrix completion based on Partially Collective Matrix Factorization.
Chao LanXiaoli LiYujie DengJoseph St. AmandJun HuanPublished in: ICPR (2016)
Keyphrases
- matrix factorization
- matrix completion
- low rank
- upper bound
- collaborative filtering
- missing data
- recommender systems
- lower bound
- nonnegative matrix factorization
- low rank matrix
- stochastic gradient descent
- negative matrix factorization
- worst case
- data matrix
- model selection
- learning algorithm
- high order
- maximum likelihood
- incomplete data
- small number
- semi supervised
- tensor factorization
- feature selection