DisCor: Corrective Feedback in Reinforcement Learning via Distribution Correction.
Aviral KumarAbhishek GuptaSergey LevinePublished in: CoRR (2020)
Keyphrases
- reinforcement learning
- uniformly distributed
- function approximation
- markov decision processes
- robotic control
- multi agent reinforcement learning
- random variables
- probability distribution
- state space
- dynamic programming
- machine learning
- multi agent
- optimal control
- spatial distribution
- action selection
- decision trees
- power law
- reinforcement learning algorithms
- learning algorithm
- policy search
- information retrieval