Advancing Transportation Mode Share Analysis with Built Environment: Deep Hybrid Models with Urban Road Network
Dingyi Zhuang, Qingyi Wang, Yunhan Zheng, Xiaotong Guo, Shenhao Wang, Haris N Koutsopoulos, Jinhua Zhao
TL;DR
This work addresses the challenge of incorporating the built environment into transportation mode-share analysis by introducing Deep Hybrid Models (DHM) that fuse graph-embedded road-network representations learned via Node2Vec with sociodemographic data. The approach uses a mixing operator to combine latent road-network features with traditional predictors and a simple predictive model (e.g., MNL) to estimate mode shares, while remaining adaptable to other data types. Empirically, the method yields about a 20% improvement in out-of-sample fit on Chicago data, with concatenated inputs achieving up to 40% gains and providing clear spatial and urban-insight explanations through GER patterns and clustering. The results demonstrate that unstructured built-environment information can substantially enhance travel-demand modeling, offering scalable, interpretable insights for planning and policy across cities and time periods.
Abstract
Transportation mode share analysis is important to various real-world transportation tasks as it helps researchers understand the travel behaviors and choices of passengers. A typical example is the prediction of communities' travel mode share by accounting for their sociodemographics like age, income, etc., and travel modes' attributes (e.g. travel cost and time). However, there exist only limited efforts in integrating the structure of the urban built environment, e.g., road networks, into the mode share models to capture the impacts of the built environment. This task usually requires manual feature engineering or prior knowledge of the urban design features. In this study, we propose deep hybrid models (DHM), which directly combine road networks and sociodemographic features as inputs for travel mode share analysis. Using graph embedding (GE) techniques, we enhance travel demand models with a more powerful representation of urban structures. In experiments of mode share prediction in Chicago, results demonstrate that DHM can provide valuable spatial insights into the sociodemographic structure, improving the performance of travel demand models in estimating different mode shares at the city level. Specifically, DHM improves the results by more than 20\% while retaining the interpretation power of the choice models, demonstrating its superiority in interpretability, prediction accuracy, and geographical insights.