Validation and Forecasts in Models Estimated from Multi-days Travel Survey

This paper will discuss how multi-day travel surveys (also known as panel data) have recently assumed high relevance in travel behavior analysis and activity-based modeling. Basically two types of multi-day data have been collected: cross-sectional data repeated at “separate” points in time, and data gathered over a “continuous” period of time. The paper shows how the studies using panel data to date have focused on the estimation of demand models, where little is known about the application and validation of models estimated on repeated measurements. Even the definition of the hold out sample is not obvious in panel data sets. This paper studies the issues related to model validation and forecasting on continuous data sets. Using both simulated and real data, this paper provides empirical evidence on the effects that different pattern of correlation have on model forecast and policy analysis. The results show that the way hold out samples are extracted affects the validation results and that the best results are obtained when using a percentage of individuals with all their observations. The paper also found that the logit model in presence of taste heterogeneity can produce biased modal shifts, while failing to account for correlation across observations does not seem to produce relevant effects on policy analysis. The real case study, estimated on a six-week travel diary (Mobidrive), confirms only in part the analysis on simulated data. The results confirm that also in panel data the model with better fit might provide worse validation and forecasting.

• Record URL:
  • https://aetransport.org/past-etc-papers/conference-papers-pre-2009/conference-papers-2010
• Corporate Authors:

  Association for European Transport (AET)

  1 Vernon Mews, Vernon Street, West Kensington
  London W14 0RL,    
• Authors:
  • Cherchi, Elisabetta
  • Cirillo, Cinzia
• Conference:
  • European Transport Conference, 2010
  • Location: Glasgow Scotland, United Kingdom
  • Date: 2010-10-11 to 2010-10-13
• Publication Date: 2010

Language

• English

Media Info

• Media Type: Digital/other
• Features: References; Tables;
• Pagination: 15p
• Monograph Title: European Transport Conference, 2010 Proceedings

Subject/Index Terms

• TRT Terms: Activity choices; Data collection; Datasets; Modal shift; Panel studies; Policy analysis; Traffic forecasting; Travel behavior; Travel demand; Travel diaries; Travel surveys; Validation
• Uncontrolled Terms: Activity based modeling
• Subject Areas: Data and Information Technology; Highways; Planning and Forecasting; Public Transportation; I72: Traffic and Transport Planning;

Filing Info

• Accession Number: 01353891
• Record Type: Publication
• Files: TRIS
• Created Date: Oct 19 2011 12:52PM