Abstract and Explore: A Novel Behavioral Metric with Cyclic Dynamics in Reinforcement Learning.
Anjie ZhuPeng-Fei ZhangRuihong QiuZetao ZhengZi HuangJie ShaoPublished in: AAAI (2024)
Keyphrases
- reinforcement learning
- markov decision processes
- dynamic model
- function approximation
- distance function
- transfer learning
- high level
- metric space
- low level
- state space
- control system
- dynamical systems
- learning problems
- search space
- evaluation metrics
- reinforcement learning algorithms
- initial conditions
- behavioral patterns
- robotic control