Neural networks and Markov models for the iterated prisoner's dilemma | IEEE Conference Publication | IEEE Xplore