Localized Speed Prediction with the use of Spatial Simultaneous Autoregressive Models

This paper examines how to employ spatial regression modelling as a direct demand modelling approach to provide localized speed estimates in a large scale network. In particular, four different spatial simultaneous autoregressive (SAR) models are estimated and compared to an ordinary linear regression in order to highlight and evaluate their capability of explaining transport related phenomena and resolving issues that arise from the underlying spatial dependence. A particular focus is given on the identification, the construction, and the selection of the spatial weighting matrices. The authors conclude that the spatial autocorrelation (SAC) model outperforms the other SAR models, resolving spatial dependence issues, and thus is the proposed one for speed prediction purposes.

• Supplemental Notes:
  • This paper was sponsored by TRB committee ADB40 Transportation Demand Forecasting.
• Corporate Authors:

  Transportation Research Board

  500 Fifth Street, NW
  Washington, DC  United States  20001
• Authors:
  • Sarlas, Georgios
  • Axhausen, Kay W
• Conference:
  • Transportation Research Board 94th Annual Meeting
  • Location: Washington DC, United States
  • Date: 2015-1-11 to 2015-1-15
• Date: 2015

Language

• English

Media Info

• Media Type: Digital/other
• Features: Figures; Maps; References; Tables;
• Pagination: 20p
• Monograph Title: TRB 94th Annual Meeting Compendium of Papers

Subject/Index Terms

• TRT Terms: Autocorrelation; Linear regression analysis; Mathematical prediction; Regression analysis; Spatial analysis; Speed; Travel demand
• Uncontrolled Terms: Autoregressive models
• Subject Areas: Planning and Forecasting; Transportation (General); I72: Traffic and Transport Planning;

Filing Info

• Accession Number: 01558336
• Record Type: Publication
• Report/Paper Numbers: 15-5536
• Files: TRIS, TRB, ATRI
• Created Date: Mar 31 2015 8:51AM