Towards Safe Reinforcement Learning via Constraining Conditional Value-at-Risk.
Chengyang YingXinning ZhouHang SuDong YanNing ChenJun ZhuPublished in: IJCAI (2022)
Keyphrases
- reinforcement learning
- function approximation
- direct policy search
- state space
- markov decision processes
- control problems
- model free
- learning algorithm
- temporal difference
- robotic control
- learning process
- machine learning
- multi agent
- reinforcement learning algorithms
- information retrieval
- relational reinforcement learning
- optimal policy
- transfer learning
- neural network
- artificial intelligence
- policy search
- partially observable domains