A Deep Learning Approach to the Prediction of Short-term Traffic Accident Risk

With the rapid development of urbanization, the boom of vehicle numbers has resulted in serious traffic accidents, which led to casualties and huge economic losses. The ability to predict the risk of traffic accident is important in the prevention of the occurrence of accidents and to reduce the damages caused by accidents in a proactive way. However, traffic accident risk prediction with high spatiotemporal resolution is difficult, mainly due to the complex traffic environment, human behavior, and lack of real-time traffic-related data. In this study, we collected heterogeneous traffic-related data, including traffic accident, traffic flow, weather condition and air pollution from the same city; proposed a deep learning model based on recurrent neural network toward a prediction of traffic accident risk. The predictive accident risk can be potential applied to the traffic accident warning system. We ranked the predictive power of various factors considered in our model through the method of Granger causality analysis, and established the order of predictive power as traffic flow > traffic accident > geographical position weather + air quality + holiday + time period, which indicate that traffic flow is the most essential factor for the occurrence of traffic accidents. The proposed method can be integrated into an intelligent traffic control system toward a more reasonable traffic prediction and command organization.