Never Stop Learning: The Effectiveness of Fine-Tuning in Robotic Reinforcement Learning.
Ryan JulianBenjamin SwansonGaurav SukhatmeSergey LevineChelsea FinnKarol HausmanPublished in: CoRL (2020)
Keyphrases
- reinforcement learning
- fine tuning
- learning process
- learning algorithm
- learning tasks
- online learning
- action selection
- supervised learning
- knowledge acquisition
- learning problems
- learning analytics
- autonomous learning
- multi agent
- prior knowledge
- temporal difference learning
- actor critic
- previously learned
- rl algorithms
- decision trees