Acquisition of rational group behavior with Mutual Reinforcement Learning

• A novel learning method for multiagent systems, named Mutual Reinforcement Learning (MRL), which applies external evaluation in multiagent systems, is proposed.
• MRL realizes acquisition of a rational group behavior based on partial observation and facilitating learning.
• MRL is applied to ad-hoc network and topologies that cover large area are learned while maintaining coverage efficiency per unit is high.