NTL Record

Title Making a Model a Good Predictive Tool
Record ID 80684
Personal Name
Creator
Rossi, Thomas
Personal Name
Contributor
Sun, Sarah
Corporate Creator United States. Federal Highway Administration
Corporate
Contributor
Cambridge Systematics, Inc.
Publisher United States. Federal Highway Administration
Publication Date 20161130
Language English
Abstract A basic part of travel demand model validation is running the model for a “base year” and comparing the outputs to observed data. Sensitivity testing and temporal validation are also critical components of any model validation effort. In this project, two model versions for each of two metropolitan areas were run twice each: 1) the more recent version for its base year and a backcast scenario (for the base year of the earlier model version), and 2) the earlier version for its base year and a forecast scenario (for the base year of the later model version). Some lessons from the comparison of the model results were presented, demonstrating the difficulty that models can have in forecasting and reinforcing the need for accuracy checking in model input data.
Rosap ID dot:55807
Rosap URL https://rosap.ntl.bts.gov/view/dot/55807
TRT Terms Travel demand; Traffic models; Traffic forecasting
General Subjects Model validation; Backcasting; Forecasting
Geographical
Coverage
United States
TRIS Online
Accession No
1648879
Contract Number DTFH61-13-F-00101
Report Number FHWA-HEP-17-048
Resource type Tech Report
URL https://ntlrepository.blob.core.windows.net/lib/80000/80600/80684/FHWA-HEP-17-048.pdf
Format PDF
Database NTL Digital Repository