1 paper · 2 filters
Alexandre Chenu, Olivier Serris, Olivier Sigaud +1
Deep Reinforcement Learning has been successfully applied to learn robotic control. However, the corresponding algorithms struggle when applied to problems where the agent is only…