发帖
雷达卡
Combining Pessimism with Optimism for Robust and Efficient
Model-Based Deep Reinforcement Learning
Sebastian Curi 1 Ilija Bogunovic 1 Andreas Krause 1
Abstract unpredictable ways. The main goal is then to learn a policy
In real-world tasks, reinforcement learning (RL) that provably brakes in a robust fashion so that, even if
agents frequently encounter situations that are not faced with new conditions, it performs reliabl ...
京公网安备 11010802022788号
