Flow to Control: Offline Reinforcement Learning with Lossless Primitive Discovery.
Yiqin YangHao HuWenzhe LiSiyuan LiJun YangQianchuan ZhaoChongjie ZhangPublished in: AAAI (2023)
Keyphrases
- reinforcement learning
- control problems
- optimal control
- control system
- knowledge discovery
- data sets
- robot control
- control strategy
- image compression
- learning process
- state space
- function approximation
- high level
- control method
- action selection
- data mining
- real time
- robotic control
- dynamic programming
- multiscale
- adaptive control
- lossless compression
- feedback control
- control policy
- scientific discovery