An actor-critic deep reinforcement learning approach for metro train scheduling with rolling stock circulation under stochastic demand

Cheng-shuo Ying, Andy H.F. Chow, Kwai‐Sang Chin