Energy-based Predictive Representations for Partially Observed Reinforcement Learning.
Tianjun ZhangTongzheng RenChenjun XiaoWenli XiaoJoseph E. GonzalezDale SchuurmansBo DaiPublished in: UAI (2023)
Keyphrases
- partially observed
- reinforcement learning
- function approximation
- learning algorithm
- state space
- multi agent
- model free
- reinforcement learning algorithms
- temporal difference
- symbolic representation
- markov decision processes
- function approximators
- higher level
- multiple representations
- predictive modeling
- learning agents
- optimal control
- action space
- action selection
- real time
- transfer learning
- learning process
- information retrieval