A Human-Like Agent Based on a Hybrid of Reinforcement and Imitation Learning | IEEE Conference Publication | IEEE Xplore