Bellman-consistent Pessimism for Offline Reinforcement Learning.
Tengyang XieChing-An ChengNan JiangPaul MineiroAlekh AgarwalPublished in: NeurIPS (2021)
Keyphrases
- reinforcement learning
- function approximation
- actor critic
- multi agent
- temporal difference learning
- reinforcement learning algorithms
- model free
- state action
- state space
- learning algorithm
- optimal policy
- markov decision processes
- dynamic programming
- linear program
- piecewise linear
- stochastic approximation
- action space
- real time
- robotic control
- temporal difference
- action selection
- learning classifier systems
- globally optimal
- supervised learning
- hidden markov models
- website
- neural network