Finding Effective Security Strategies through Reinforcement Learning and Self-Play | IEEE Conference Publication | IEEE Xplore