Reinforcement Learning With Reward Machines in Stochastic Games

TL;DR

This paper introduces Q-learning with reward machines for stochastic games, enabling multi-agent systems to learn Nash equilibrium strategies in complex, non-Markovian reward environments with proven convergence properties.

Contribution

It develops a novel algorithm, QRM-SG, that incorporates reward machines into multi-agent reinforcement learning for stochastic games, with convergence guarantees to Nash equilibrium.

Findings

QRM-SG effectively learns best-response strategies in complex stochastic games.

QRM-SG converges faster than baseline methods like Nash Q-learning and MADDPG.

The algorithm demonstrates successful convergence in three case studies.

Abstract

We investigate multi-agent reinforcement learning for stochastic games with complex tasks, where the reward functions are non-Markovian. We utilize reward machines to incorporate high-level knowledge of complex tasks. We develop an algorithm called Q-learning with reward machines for stochastic games (QRM-SG), to learn the best-response strategy at Nash equilibrium for each agent. In QRM-SG, we define the Q-function at a Nash equilibrium in augmented state space. The augmented state space integrates the state of the stochastic game and the state of reward machines. Each agent learns the Q-functions of all agents in the system. We prove that Q-functions learned in QRM-SG converge to the Q-functions at a Nash equilibrium if the stage game at each time step during learning has a global optimum point or a saddle point, and the agents update Q-functions based on the best-response strategy at this point. We use the Lemke-Howson method to derive the best-response strategy given current Q-functions. The three case studies show that QRM-SG can learn the best-response strategies effectively. QRM-SG learns the best-response strategies after around 7500 episodes in Case Study I, 1000 episodes in Case Study II, and 1500 episodes in Case Study III, while baseline methods such as Nash Q-learning and MADDPG fail to converge to the Nash equilibrium in all three case studies.