Congratulations to Dr. Haotian Wang for “satisfying the requirements for the award of the degree of Doctor of Philosophy at the University of Sydney.”
Thesis Title: "An Ensemble Approach to Route Choice”
Lead Supervisor: Professor David Levinson.
Co-Supervisor: Dr. Emily Moylan.
Abstract:
Understanding automobile drivers’ route preferences allows for the calculation of traffic flow on network segments and helps in assessing facility requirements, costs, and the impact of network modifications. This can be seen as a two-stage problem. For a large urban network, available routes from an origin to a destination, which are difficult to itemize, could be numerous. Therefore, as the first step, a clear choice set of available routes for a trip should be extracted from the network before attempting to predict choice of route. The second step is predicting individuals’ route choices, which starts at the individual level, whereas traffic assignment begins at the network level and aims to determine how many travelers use each link in the network. Much previous research employs logit-based choice methods to model the route choices of individuals, but machine learning models are gaining increasing interest. However, all of these methods typically rely on a single ‘best’ model for predictions, which may be sensitive to measurement errors in the training data. Moreover, predictions from discarded models might still provide insights into route choices. The ensemble approach combines outcomes from multiple individual models using various pattern recognition methods, assumptions, and/or data sets to deliver improved predictions. When configured correctly, ensembles offer greater prediction accuracy and better account for uncertainties.
To examine the advantages of ensemble techniques, a hybrid method, which combines labelling and link-penalty approaches, is applied to a high resolution road network to prepare the choice set for route choice modelling. A data set from the I-35W Bridge Collapse study in 2008, and another from the 2011 Travel Behavior Inventory (TBI), both in Minneapolis - St. Paul (The Twin Cities) are each used to train a set of route choice models. These models are combined with ensemble techniques. The analyses consider travellers’ socio-demographics and trip attributes.
The trained models are applied to two datasets, the Longitudinal Employer-Household Dynamics (LEHD) commute trips and TBI morning peak trips, for validation. Predictions are also compared with the loop detector records on freeway links. Traditional Multinomial Logit and Path-size Logit models, along with machine learning methods such as Decision Tree, Random Forest, Extra Tree, AdaBoost, Support Vector Machine, and Neural Network, serve as the foundation for this study. Ensemble rules are tested in both case studies, including hard voting, soft voting, ranked choice voting, and stacking.
Based on the results, using the 10 best labels identified in the study, the choice set captures most observed trajectories without needing to remove any links from the road network. A new similarity measure, which considers the influence of overlap, attribute similarity, and spatial similarity between routes, is proposed and applied to evaluate route choice models. We conclude that ensembles, when properly applied, perform better than base models in route choice prediction in the datasets in the study. Additionally, heterogeneous ensembles using soft voting outperform both the base models and other ensemble rules on testing sets, with unscaled weights in soft voting proving to be more robust based on external validation.
Related Publications:
Wang, Haotian, Moylan, E. and Levinson, D. (2023) Ensemble Methods for Route Choice . Transportation Research part C. Volume 167, October 2024, 104803 [doi]
Wang, Haotian, Moylan, E. and Levinson, D. (2023) Route Choice Set Generation on High Resolution Networks . Transportation Research Record [doi]
Wang, Haotian, Moylan, E. and Levinson, D. (2022) Prediction of the Deviation between Alternative Routes and Actual Trajectories for Bicyclists. Findings, June. [doi].
