Reinforcement Learning to Efficiently Recover Control Performance of Robots Using Imitation Learning After Failure.
Shoki KobayashiTakeshi ShibuyaPublished in: SMC (2022)
Keyphrases
- imitation learning
- reinforcement learning
- robot behavior
- control problems
- robotic systems
- robot control
- reinforcement learning methods
- optimal control
- adaptive control
- control strategies
- humanoid robot
- function approximation
- autonomous robots
- state space
- learning problems
- reinforcement learning algorithms
- learning algorithm
- machine learning
- action selection
- transfer learning
- maximum margin
- dynamic programming
- multi agent
- control method
- markov decision processes
- control system