Cross-domain adaptive transfer reinforcement learning based on state-action correspondence.
Heng YouTianpei YangYan ZhengJianye HaoMatthew E. TaylorPublished in: UAI (2022)
Keyphrases
- cross domain
- state action
- reinforcement learning
- transfer learning
- knowledge transfer
- evaluation function
- action space
- function approximators
- average reward
- function approximation
- learning tasks
- stochastic games
- text categorization
- markov decision process
- temporal difference
- cross domain learning
- state transitions
- machine learning
- markov decision processes
- state space
- multi agent
- reinforcement learning algorithms
- optimal policy
- semi supervised learning
- target domain
- model free
- action selection
- e government
- learning capabilities
- decision making
- real valued
- e learning
- learning process
- active learning
- collaborative filtering
- text mining
- reward function
- optimal control
- distance metric