Ensemble Methods For Route Choice

Route selection involves determining the paths between origins and destinations within a network. Understanding travellers’ 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. Traditionally, statistical models are employed for route choice modelling, 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, discarded models might still provide insights into route choices. Ensemble modelling combines outcomes from models using various pattern recognition methods, assumptions, or data sets to deliver improved predictions. When configured correctly, ensemble models offer greater prediction accuracy and account for uncertainties. To examine the advantages of ensemble techniques in route choice modelling, a data set from the I-35W Bridge Collapse study in 2008, and another from 2011 Travel Behavior Inventory in Minneapolis-St. Paul (The Twin Cities) are used. The analysis considered travellers’ socio-demographics and trip attributes. 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, served as the foundation for this study. Ensemble rules were tested in both case studies, including hard voting, soft voting, ranked choice voting, weighted soft voting, and stacking. Based on the results, ensemble models using weighted soft voting outperform other ensemble rules and generally have the highest ranking in sensitivity and precision.