Sample Complexity of Decentralized Tabular Q-Learning for Stochastic Games.
Zuguang GaoQianqian MaTamer BasarJohn R. BirgePublished in: ACC (2023)
Keyphrases
- sample complexity
- stochastic games
- multi agent
- reinforcement learning algorithms
- multiagent reinforcement learning
- learning algorithm
- single agent
- reinforcement learning
- cooperative
- learning agent
- theoretical analysis
- learning problems
- upper bound
- special case
- markov decision processes
- generalization error
- supervised learning
- active learning
- model free
- state space
- lower bound
- optimal policy
- nash equilibria
- sample size
- training examples
- function approximation
- multi agent systems
- average reward
- infinite horizon
- sequential decision problems
- machine learning algorithms
- semi supervised learning
- reward function
- policy iteration
- model selection
- semi supervised
- feature selection