In this demo, we will present a simulation-based human-computer interaction of deep reinforcement learning in action on order dispatching and driver repositioning for ride-sharing. Specifically, we will demonstrate through several specially designed domains how we use deep reinforcement learning to train agents (drivers) to have longer optimization horizon and to cooperate to achieve higher objective values collectively.
Modelling, Detecting and Mitigating Threats against Industrial Healthcare Systems: A combined SDN and Reinforcement Learning Approach