DeepTransport: Prediction and Simulation of Human Mobility and Transportation Mode at a Citywide Level / 2618
Xuan Song, Hiroshi Kanasugi, Ryosuke Shibasaki
Traffic congestion causes huge economic loss worldwide in every year due to wasted fuel, excessive air pollution, lost time, and reduced productivity. Understanding how humans move and select the transportation mode throughout a large-scale transportation network is vital for urban congestion prediction and transportation scheduling. In this study, we collect big and heterogeneous data (e.g., GPS records and transportation network data), and we build an intelligent system, namely DeepTransport, for simulating and predicting human mobility and transportation mode at a citywide level. The key component of DeepTransport is based on the deep learning architecture that that aims to understand human mobility and transportation patterns from big and heterogeneous data. Based on the learning model, given any time period, specific location of the city or people's observed movements, our system can automatically simulate or predict the persons' future movements and their transportation mode in the large-scale transportation network. Experimental results and validations demonstrate the efficiency and superior performance of our system, and suggest that human transportation mode may be predicted and simulated more easily than previously thought.