Skip to content

Optimizing the initialization of fleet deployment for mobility-on-demand systems: A deep reinforcement learning approach