Expert-Supervised Reinforcement Learning for Offline Policy Learning and Evaluation.
Aaron Sonabend W.Junwei LuLeo Anthony CeliTianxi CaiPeter SzolovitsPublished in: NeurIPS (2020)
Keyphrases
- reinforcement learning
- learning algorithm
- supervised learning
- learning process
- evolutionary learning
- optimal policy
- learning problems
- policy search
- partially observable
- learning tasks
- learning systems
- machine learning
- learning capabilities
- action selection
- partially observable environments
- markov decision process
- human experts
- inverse reinforcement learning
- knowledge acquisition
- state space
- dynamic programming
- active learning
- markov decision processes
- unsupervised learning
- mobile robot
- rl algorithms
- policy gradient
- prior knowledge
- eligibility traces
- real time