Detecting Rewards Deterioration in Episodic Reinforcement Learning
Ido Greenberg 1 Shie Mannor 1 2
Abstract RL tasks is the safety and reliability of the system (Dulac-
Arnold et al., 2019; Chan et al., 2020), arising in both of-
In many RL applications, once training ends, it is fline and online settings.
vital to detect any deterioration in the agent per-
formance as soon as possible. Furthermore, i ...


雷达卡




京公网安备 11010802022788号







