Multiagent Reinforcement Learning-Based Taxi Predispatching Model to Balance Taxi Supply and Demand

Figure 1

Processing pipeline.