RESEARCH ARTICLE
Development of Improved Models for Imputing Missing Traffic Counts
Ming Zhong*, 1, Satish Sharma2
Article Information
Identifiers and Pagination:
Year: 2009Volume: 3
First Page: 35
Last Page: 48
Publisher ID: TOTJ-3-35
DOI: 10.2174/1874447800903010035
Article History:
Received Date: 10/10/2008Revision Received Date: 23/12/2008
Acceptance Date: 25/12/2008
Electronic publication date: 9/3/2009
Collection year: 2009
open-access license: This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 International Public License (CC-BY 4.0), a copy of which is available at: https://creativecommons.org/licenses/by/4.0/legalcode. This license permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Abstract
Estimating missing values is known as data imputation. A literature review indicates that many highway and transportation agencies in North America and Europe use various traditional methods to impute their traffic counts. These methods can be broadly categorized into factor and time series analysis approaches. However, little or no research has been conducted to assess the imputation accuracy. The literature indicates that the current practices are varied, and the methods used by highway agencies are intuitive in nature. Typical imputation methods used by highway agencies are identified and applied to data from six automatic traffic recorders (ATRs) in Alberta, Canada, to evaluate their accuracy. Statistical analysis shows that these traditional methods result in varying accuracy for traffic counts from different types of roads. In some cases, the imputation errors can be unacceptably high. Therefore, improved imputation methods are proposed. Study results indicate that imputation accuracy can be significantly improved by incorporating correction factors and data from both before and after the failure periods into the traditional models. The improved imputations should provide transportation practitioners better information for decision marking purposes.
