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

Zihang Wang, Maowei Jiang

Reinforcement learning(RL) algorithms face the challenge of limited data efficiency, particularly when dealing with high-dimensional state spaces and large-scale problems. Most of…