This study aims to explore the transportation mode choice of passengers in intercity travel, with a particular focus on applying machine learning techniques to analyze the key factors influencing travel decisions. This study is based on mobile phone signaling data from 41 cities in the Yangtze River Delta region in 2019, and constructs an intercity travel table for intercity trips between cities in the Yangtze River Delta and Shanghai, which serves as the subject of this research. Various features, including travel distance, travel time, and travel speed, are extracted to build machine learning models for analyzing their impact on transportation mode choice. In terms of model construction, this study compares the performance of logistic regression, random forests, and XGBoost, finding that XGBoost outperforms the others in terms of prediction accuracy and model efficiency. Through the analysis of SHAP values and partial dependence plots, the study further reveals the underlying mechanisms of the influence of various factors. For example, the time spent in the destination city significantly impacts the choice of transportation mode, while travel distance and speed exhibit varying levels of appeal for short, medium, and long-distance travellers. The research also identifies significant differences in transportation preferences among different demographic groups. For instance, for short-distance travel, younger passengers tend to prefer railway travel whereas middle-aged and older individuals are more likely to choose highway travel. The findings provide important insights for optimizing intercity transportation planning and enhancing service quality. Based on the needs of different passenger groups, recommendations are made, such as increasing train frequency and optimizing night time transportation services. The study also acknowledges its limitations, including the failure to account for external factors such as weather conditions and travel purposes. Future research will aim to incorporate additional data sources and more complex scenarios to further enhance the model’s predictive capability and practical application value.
Identification and Influencing Factors Analysis for Intercity Travel Modes Based on Mobile Phone Signaling Data
Lect. Notes Electrical Eng.
International Symposium for Intelligent Transportation and Smart City ; 2025 ; Shanghai, China May 16, 2025 - May 17, 2025
International Symposium for Intelligent Transportation and Smart City (ITASC) 2025 Proceedings ; Kapitel : 18 ; 165-177
20.07.2025
13 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Europäisches Patentamt | 2023
|British Library Conference Proceedings | 2023
|Intercity transportation network feature analysis method and device based on mobile phone signaling
Europäisches Patentamt | 2025
|NTIS | 1975
|