Nearly Optimal Control for Mixed Zero-Sum Game Based on Off-Policy Integral Reinforcement Learning | IEEE Journals & Magazine | IEEE Xplore