Actively Learning Costly Reward Functions for Reinforcement Learning.
André EberhardHoussam MetniGeorg FahlandAlexander StrohPascal FriederichPublished in: CoRR (2022)
Keyphrases
- reinforcement learning
- learning algorithm
- learning process
- reward function
- inverse reinforcement learning
- state space
- function approximation
- active learning
- state action
- reinforcement learning algorithms
- optimal policy
- reinforcement learning methods
- learning agents
- learning tasks
- supervised learning
- probabilistic model
- machine learning
- information extraction
- policy search
- prior knowledge
- hierarchical reinforcement learning