A Study on Cooperative Action Selection Considering Unfairness in Decentralized Multiagent Reinforcement Learning

Reinforcement learning has been studied for cooperative learning and optimization methods in multiagent systems. In several frameworks of multiagent reinforcement learning, the system's whole problem is decomposed into local problems for agents. To choose an appropriate cooperative action, the agents perform an optimization method that can be performed in a distributed manner. While the conventional goal of the learning is the maximization of the total rewards among agents, in practical resource allocation problems, unfairness among agents is critical. In several recent studies of decentralized optimization methods, unfairness was considered a criterion. We address an action selection method based on leximin criteria, which reduces the unfairness among agents, in decentralized reinforcement learning. We experimentally evaluated the effects and influences of the proposed approach on classes of sensor network problems.