Deep Reinforcement Learning from Self-Play in Imperfect-Information Games

Johannes Heinrich, David Silver

Many real-world applications can be described as large-scale games of imperfect information. To deal with these challenging domains, prior work has focused on computing Nash equili…