Safe Exploration for Model-based Reinforcement Learning using Optimistic Constraints
Published in Artificial Intelligence and Statistics 2023 (under review), 2023
This paper tackles the problem of Safe Reinforcement Learning using uncertainty-aware classifier to construct optimistic yet safe plans.
Recommended citation: Nikita Kostin, Aidan Scannell, Arno Solin, Joni Pajarinen (2023)