Exploring the relationship of reward and punishment in reinforcement learning | IEEE Conference Publication | IEEE Xplore