Data-Based Optimal Control of Multiagent Systems: A Reinforcement Learning Design Approach | Synapse