S4RL: Surprisingly Simple Self-Supervision for Offline Reinforcement Learning.
Samarth SinhaAnimesh GargPublished in: CoRR (2021)
Keyphrases
- reinforcement learning
- function approximation
- multi agent
- state space
- optimal policy
- model free
- reinforcement learning algorithms
- approximate dynamic programming
- active learning
- temporal difference
- action selection
- neural network
- control problems
- transfer learning
- supervised learning
- real time
- markov decision processes
- learning classifier systems
- learning capabilities
- partially observable
- machine learning
- action space
- control policy
- learning agents
- autonomous learning
- actor critic
- partially observable domains