Nearly Minimax Optimal Offline Reinforcement Learning with Linear Function Approximation: Single-Agent MDP and Markov Game.
Wei XiongHan ZhongChengshuai ShiCong ShenLiwei WangTong ZhangPublished in: CoRR (2022)
Keyphrases
- reinforcement learning
- function approximation
- single agent
- temporal difference learning algorithms
- stochastic games
- multi agent
- function approximators
- markov decision processes
- optimal control
- action space
- average reward
- dynamic programming
- optimal policy
- markov decision process
- state space
- temporal difference
- reinforcement learning algorithms
- model free
- policy gradient
- control policies
- temporal difference learning
- control policy
- approximate dynamic programming
- markov decision problems
- game tree
- multiple agents
- supervised learning
- learning process
- partially observable
- learning agent
- learning algorithm
- radial basis function
- markov chain
- decision problems
- policy iteration
- dynamic environments
- optimal solution
- reinforcement learning problems
- multi agent systems
- action selection
- reinforcement learning methods
- state action
- data mining
- nash equilibrium
- domain independent
- machine learning