An Optimistic Perspective on Offline Reinforcement Learning.
Rishabh AgarwalDale SchuurmansMohammad NorouziPublished in: ICML (2020)
Keyphrases
- reinforcement learning
- function approximation
- optimal policy
- markov decision processes
- learning classifier systems
- machine learning
- viewpoint
- temporal difference
- state space
- model free
- real time
- supervised learning
- direct policy search
- multi agent reinforcement learning
- control problems
- action selection
- transfer learning
- least squares
- dynamic programming
- learning process
- information systems
- learning algorithm
- neural network