Graphical Minimax Game and On-Policy Reinforcement Learning for Consensus of Leaderless Multi-Agent Systems | IEEE Conference Publication | IEEE Xplore