LILAC: Learning a Leader for Cooperative Reinforcement Learning | IEEE Conference Publication | IEEE Xplore