An Improved Matrix Approximation for Recommender Systems Based on Context-Information and Transfer Learning.
Man LiLuosheng WenPublished in: IEEE Access (2019)
Keyphrases
- transfer learning
- recommender systems
- collaborative filtering
- matrix approximation
- least squares
- learning tasks
- approximation error
- knowledge transfer
- cross domain
- labeled data
- matrix factorization
- theoretical guarantees
- low rank matrix approximation
- multi task
- maximum entropy
- reinforcement learning
- structure learning
- machine learning
- transfer knowledge
- text categorization
- semi supervised learning
- active learning
- text mining
- machine learning algorithms
- text classification
- learning problems
- learning algorithm
- target domain
- training data
- learning process