Investigation of Effects of Inherent Variation and Spatiotemporal Dependency on Urban Travel-Speed Prediction

Abstract

Urban traffic prediction is a challenging task due to the complexity of urban networks. Many studies have been conducted to improve the prediction accuracy, but the limitation still remains that their accuracy varies with location and time due to lack of understanding. To overcome this limitation, it is necessary to investigate in depth the various phenomena that change the traffic flow patterns. Among the phenomena, this study aims to analyze the effect of inherent variation in a link and spatiotemporal dependency between links in predicting travel speed in urban networks and to identify the factors that influence the two phenomena. The results show that the variation and dependency have significant differences according to locations. The results also indicate that the effects of the two phenomena vary depending on the prediction horizon of the prediction model and suggest to consider both the variation and dependency in short-term prediction but focus on only the variation in long-term prediction. The authors also identify the factors that affect the two phenomena and recommend guidelines for urban traffic prediction.