When are Offline Two-Player Zero-Sum Markov Games Solvable?

Qiwen Cui Simon S. Du

Published in: NeurIPS (2022)

Keyphrases

markov games
markov decision processes
reinforcement learning algorithms
multiagent reinforcement learning
reinforcement learning
markov decision process
control problems

special case
np hard
state space
multiagent systems
computational complexity
model free
nash equilibrium

multi agent
finite state
temporal difference learning
cooperative
infinite horizon
function approximation
optimal policy

adaptive control
temporal difference
convergence rate
dynamic environments
stochastic games
cost function