Deep Learning for Control: a non-Reinforcement Learning View | IEEE Conference Publication | IEEE Xplore