Nonlinear Relationship between Built Environments and Metro Ridership at Station-to-Station Level on Machine Learning Methods: A Comparison of Commuters and Non-Commuters

ABSTRACTTo reveal the non-linear relationship between built environments and station-to-station ridership of commuters, this study firstly identified the urban residents’ rail commuters and non-commuter users using K-means clustering algorithm. Gradient Boosting Decision Tree (GBDT) was then used to predict the origin and destination station-to-station ridership. Using the automatic fare collection data and POI data of Xi’an, the models are evaluated. The results show that the importance of the number of bus stops and the road density around the metro station is high for commuters, while the number of leisure and entertainment destinations around the metro station and the number of parking lots are more important for non-commuters.


  • English

Media Info

  • Pagination: pp 2573-2583
  • Monograph Title: CICTP 2023: Innovation-Empowered Technology for Sustainable, Intelligent, Decarbonized, and Connected Transportation

Subject/Index Terms

Filing Info

  • Accession Number: 01906374
  • Record Type: Publication
  • ISBN: 9780784484869
  • Files: TRIS, ASCE
  • Created Date: Jan 30 2024 9:23AM