D4RL: Datasets for Deep Data-Driven Reinforcement Learning.
Justin FuAviral KumarOfir NachumGeorge TuckerSergey LevinePublished in: CoRR (2020)
Keyphrases
- reinforcement learning
- data driven
- reinforcement learning algorithms
- function approximation
- state space
- markov decision processes
- multi agent
- control problems
- action selection
- model free
- learning algorithm
- temporal difference
- optimal policy
- machine learning
- rl algorithms
- optimal control
- direct policy search
- transfer learning
- policy iteration
- temporal difference learning
- learning problems
- supervised learning
- multi agent reinforcement learning
- state action
- approximate dynamic programming
- policy evaluation
- continuous state
- learned knowledge
- function approximators
- synthetic datasets