An Overview and General Framework for Spatiotemporal Modeling and Applications in Transportation and Public Health

Abstract

Spatiotemporal modeling and forecasting is an essential task for many real-world problems, especially in the field of transportation and public health. The complex and dynamic patterns with dual attributes of time and space create unique challenges for effective modeling and forecasting. With the advancement of data collection, storage, and sharing technologies, the amount of data and the types of data available for spatiotemporal modeling research in transportation and public health are rapidly increasing. Some traditional spatiotemporal methods become obsolete. There is a need to review existing methods and propose new ones to harness the power of newly available data. Therefore, in this chapter, we conduct a comprehensive survey of methods and algorithms for spatiotemporal monitoring and forecasting, focusing on applications in transportation and public health. Then, we propose a systematic framework to incorporate three different approaches: statistical methods, machine learning methods, and mechanistic simulation methods. The proposed framework is expected to help researchers in the field to better formulate spatiotemporal problems, construct appropriate models, and facilitate new developments that combine the strengths of mechanistic approaches and data-driven ones. The proposed general framework is illustrated via examples of spatiotemporal methods developed in transportation and public health.