Bisimulation metric for Model Predictive Control

Yutaka Shimizu, Masayoshi Tomizuka

Model-based reinforcement learning has shown promise for improving sample efficiency and decision-making in complex environments. However, existing methods face challenges in train…