https://deepai.org/publication/computably-continuous-reinforcement-learning-objectives-are-pac-learnable
Computably Continuous Reinforcement-Learning Objectives are PAC-learnable | DeepAI
Mar 9, 2023 - 03/09/23 - In reinforcement learning, the classic objectives of maximizing discounted and finite-horizon cumulative rewards are PAC-learnable...
reinforcement learningcontinuousobjectivespaclearnable