Model-Based Multi-Agent RL in Zero-Sum Markov Games with Near-Optimal Sample Complexity.
Kaiqing ZhangSham M. KakadeTamer BasarLin F. YangPublished in: NeurIPS (2020)
Keyphrases
- markov games
- sample complexity
- multiagent reinforcement learning
- multi agent
- reinforcement learning
- reinforcement learning algorithms
- markov decision processes
- model free
- learning problems
- learning algorithm
- supervised learning
- markov decision process
- theoretical analysis
- stochastic games
- multiagent systems
- upper bound
- cooperative
- active learning
- lower bound
- special case
- control problems
- generalization error
- policy iteration
- state space
- training examples
- multi agent systems
- function approximation
- finite state
- autonomous agents
- dynamic programming
- sample size
- optimal control
- support vector
- infinite horizon
- temporal difference
- transfer learning
- single agent
- optimal policy
- learning agent
- dynamic environments