Enhancing data efficiency in reinforcement learning: a novel imagination mechanism based on mesh information propagation

Open in new window