Pre-Training for Robots: Offline RL Enables Learning New Tasks from a Handful of Trials.
Aviral KumarAnikait SinghFrederik EbertYanlai YangChelsea FinnSergey LevinePublished in: CoRR (2022)
Keyphrases
- reinforcement learning
- learning process
- learning algorithm
- supervised learning
- learning tasks
- autonomous robots
- learning systems
- multiple tasks
- multi task
- learning problems
- online learning
- active learning
- transfer learning
- multi task learning
- learning speed
- training set
- robot control
- autonomous learning
- solve complex tasks
- robotic systems
- learning experience
- learning styles
- semi supervised learning
- knowledge acquisition
- state space
- multi agent
- e learning