Deep Reinforcement Q-Learning for Security Games
The project is to use advances in Reinforcement learning to learn the probabilities of protecting an assest to optimize the defenders maximum reward while minimizing the attackers maximum reward
The project is web-based and run purely through javascript.