THE USE OF PRIOR INFORMATION IN GRAVITY MODEL CALIBRATION

The calibration of a gravity model with an empirical deterrence function can sometimes lead into difficulties of non-identifiability of some of the parameters when partial matrix methods are used. The essential character of the problem is described and is shown to be similar to two other problems: estimating the turning flows at an intersection from approach road counts and determining the most likely trip matrix from link volumes. In all three problems prior information is required to recover full identifiability. It is argued that the most appropriate way in which to use the combination of prior beliefs and observations is Bayesian inference, and a simple modification to the calibration process is proposed. A comparison is made of the principles of maximum entropy and minimum information with the Bayesian approach. (Author/TRRL)

• Availability:
  • Find a library where document is available. Order URL: http://worldcat.org/issn/00410683
• Corporate Authors:

  Printerhall Limited

  29 Newmart Street
  London W1P 3PE,   England 
• Authors:
  • MAHER, M J
• Publication Date: 1983-2

Media Info

• Features: References; Tables;
• Pagination: p. 68
• Serial:
  • Traffic Engineering & Control
  • Volume: 24
  • Issue Number: 2
  • Publisher: Hemming Group, Limited
  • ISSN: 0041-0683
  • Serial URL: http://www.tecmagazine.com/

Subject/Index Terms

• TRT Terms: Bayes' theorem; Calibration; Costs; Distributions (Statistics); Gravity models; Intersections; Link volume; Mathematical models; Matrices (Mathematics); Traffic flow; Travel; Turning traffic
• Uncontrolled Terms: Turning movements
• ITRD Terms: 6171: Calibration; 224: Cost; 6572: Distribution (stat); 690: Gravity model; 698: Journey; 6473: Mathematical model; 6494: Matrix
• Subject Areas: Finance; Highways; Operations and Traffic Management; I71: Traffic Theory;

Filing Info

• Accession Number: 00373979
• Record Type: Publication
• Source Agency: Transport Research Laboratory
• Report/Paper Numbers: HS-034 632
• Files: ITRD, TRIS
• Created Date: Jul 30 1983 12:00AM