Improving the approaches of traffic demand forecasting in the big data era.

Abstract Since the 2000s, an era of big data has emerged. Since then, urban planners have increasingly applied the theory and methods of big data in planning practice. Recent decades illustrate a rapid increase of the application of big data approaches in transportation, bringing new opportunities for innovation in transport modeling. This article analyzes the theories and methods of big data in traffic demand forecasting. In view of theory, the new models and algorithms are proposed in order to adapt to new big data and response to the limitations of traditional disaggregated approaches. In such approaches, three traffic demand-forecasting methods, the full sample-demand distribution model, the traffic integration model, the model organism protein expression database model, are discussed. Undoubtedly, the development of big data also presents new challenges to travel-demand forecasting methods regarding data acquisition, data processing, data analysis, and application of results. In particular, identifying how to improve approaches to traffic-demand forecasting in the big data era in the Third World will be a challenge to the researchers in the field. Highlights • Big data brings new opportunities for the innovation of transport modeling • The new models and algorithms are proposed responding to the limitations of the traditional disaggregated approaches. • Full sample demand distribution model, Traffic integration model, and MoPeD model are discussed. • Improvement of traffic demand forecasting approaches in the big data era in the Third World will be a challenge. [ABSTRACT FROM AUTHOR]