Decentralized multi-task reinforcement learning policy gradient method with momentum over networks.
Shi JunruWang QiongMuhua LiuZhihang JiRuijuan ZhengQingtao WuPublished in: Appl. Intell. (2023)
Keyphrases
- gradient method
- multi task
- actor critic
- policy gradient
- reinforcement learning
- multi task learning
- learning tasks
- convergence rate
- transfer learning
- learning problems
- step size
- optimal policy
- learning rate
- negative matrix factorization
- multi class
- feature selection
- optimization methods
- learning algorithm
- collaborative filtering
- supervised learning
- machine learning
- machine learning algorithms
- data sets