Rule-based versus utility-maximizing models of activity-travel patterns

This paper describes the results of a comparative analysis of three activity-based travel demand models. Considered are ALBATROSS, a rule-based model, and PCA TS and LLPM -both sequential/nested logit models. The activity diary data used was split into an estimation and an holdout set. The models are estimated independently of each other on the estimation set. Various measures of goodness-of-fit are used to assess the ability of the models to predict the observed activity-travel patterns of the holdout set. The paper discusses the measures of goodness-of-fit and the results of the comparison. (a) For the covering entry of this conference, please see ITRD abstract no. E213535.

• Authors:
  • ARENTZE, T
  • BORGERS, A
  • HOFMAN, F
  • FUJII, S
  • JOH, C
  • KIKUCHI, A
  • KITAMURA, R
  • TIMMERMANS, H
  • VAN DER WAERDEN, P
• Publication Date: 2000

Language

• English

Media Info

• Pagination: 22P
• Monograph Title: Two-step estimation of discrete/continuous econometric models with interdependent multinomial choices
• Serial:
  • INTERNATIONAL CONFERENCE ON TRAVEL BEHAVIOUR RESEARCH, 9TH, 2000, GOLD COAST, QUEENSLAND, AUSTRALIA, VOL 3
  • Publisher: University of Sydney

Subject/Index Terms

• TRT Terms: Behavior; Conferences; Data collection; Mathematical analysis; Mathematical models; Scale models; Travel
• ITRD Terms: 6471: Analysis (math); 9001: Behaviour; 8525: Conference; 8623: Data acquisition; 698: Journey; 6473: Mathematical model; 6205: Model (not math)
• Subject Areas: Bridges and other structures; Operations and Traffic Management; Planning and Forecasting; I71: Traffic Theory; I72: Traffic and Transport Planning;

Filing Info

• Accession Number: 01032089
• Record Type: Publication
• Source Agency: ARRB
• Files: ITRD, ATRI
• Created Date: Sep 6 2006 1:14PM