Bias Modeling and Estimation of Networked Transportation Data

This proposed research aims to develop theories, models, and algorithms for data bias modeling and estimation on transportation networks. The proposed new NETwork-based, Data-Assisted Transportation Analysis (NetData) framework models and estimates data bias by recognizing and utilizing underlying network structures and processes in the data. The major research objectives are:

  1. Construct fundamental concepts, such as data representativeness, which provides a comprehensive performance measure of data bias estimation.
  2. Develop data taxonomy and analysis methods to identify the existence and types of bias in real world data.
  3. Investigate scientific bias estimation methods that specifically address challenges imposed by networked data. This includes the NetData bias estimation framework that integrates various network models with multiple data sources.
  4. Extend the NetData models to capture more realistic scenarios such as dynamic networks and multi-modal networks.
  5. Test and validate the theories, models, and algorithms of the proposed bias estimation methods under hypothetical and realistic settings.

The main tasks of the proposed research are:

  • Task 1 (T1) conducts in-depth data categorization and analysis of data bias issues, and develops data representativeness (DR) measures and specific data bias modeling techniques. This addresses the above first and second research objectives.
  • Task 2 (T2) develops the NetData framework, including deterministic and stochastic NetData models, which models and estimates data bias, and integrates datasets with proper network models. T2 also investigates computational methods to solve large-scale NetData models. It addresses the third research objective.
  • Task 3 (T3) extends the NetData models to capture more realistic scenarios such as dynamic networks and multi-modal networks. This address the fourth research objective of the project.
  • Task 4 (T4) conducts model testing and validation, which address the fifth research objective.

Research Highlights (TODO)

List of Publications

Journal Papers (Published, Accepted, or Submitted)

  1. Guo, Q.*, Ban, X., Aziz, H.M.A., 2021. Mixed traffic flow of human driven vehicles and connected/automated vehicles on a dynamic transportation network. Transportation Research Part C 128, 103159. The paper is also accepted for presentation at the International Symposium on Transportation and Traffic Theory (ISTTT), 2022.
  2. Wang, J.*, Liu, H., Lu, S., Ban, X., 2021. Insignificant-based origin-destination demand estimation method. Submitted to Transportation Science (2nd revision).

Working Papers

  1. Wang, J.*, Zhao, C., Ban, X., 2022. Distributionally Robust Insignificance-based Origin DestinationDemand Estimation. In preparation.
  2. Zhang, Y.*, Ban, X., 2022. Bias modeling and mitigation in app-based transportation big data. In preparation.

Conference Papers

  1. Wang, J.*, Ban, X., 2022. Transportation origin-destination demand estimation with quasi-sparsity. Presented at the ASCE-ICTD conference, Seattle, WA, June 02, 2022.
  2. Wang, J.*, Lu, S., Ban, X., 2020. Exploring Insignificant OD Pairs: A Compressed Sensing Model for OD Demand Estimation, Presented at the 99th Annual Meeting of Transportation Research Board, Washington, DC.

Invited Seminars

  1. School of Information Engineering, Chang An University, China, Network Traffic Control with CAVs. April 2021 (Webinar).
  2. School of Civil and Environmental Engineering, Georgia Tech, Macroscopic dynamic traffic control with drivers’ route choice behavior. February 2021 (webinar).
  3. Department of Industrial & Systems Engineering, University of Southern California, Macroscopic dynamic traffic control with drivers’ route choice behavior. February 2021 (webinar).
  4. Washington Department of Transportation (WSDOT), Transportation big data: promises, issues, and potential solutions, March, 2020.

Data Sets

  1. Vehicle trajectory data from Downtown Seattle (generated by simulation): Dataset 1; Dataset 2
  2. Aggregated mobility data and computer codes for analyzing the impacts of SR-99 Tunnel in Seattle: GitHub

PI: Jeff Ban (UW); Collaborator: Yueyue Fan (UC Davis)

Sponsor: National Science Foundation

Close Menu