Reinforcement Learning under Threats

Victor Gallego, Roi Naveiro, David Rios Insua

In several reinforcement learning (RL) scenarios, mainly in security settings, there may be adversaries trying to interfere with the reward generating process. In this paper, we in…