Predicting freeway incident duration using machine learning
5 affiliations
- 1Key Laboratory of Intelligent Policing and National Security Risk Management, Sichuan Police College, Luzhou, China
- 2Department of Road Traffic Management, Sichuan Police College, Luzhou, China
- 3Intelligent Policing Key Laboratory of Sichuan Province, Luzhou, China
- 4College of Automobile and Traffic Engineering, Nanjing Forestry University, Nanjing, Jiangsu, China
- 5Department of Information Systems and Business Analytics, Kent State University, Ohio, United States of America
Abstract
Accurate prediction of freeway incident duration is important for traffic management and emergency resource allocation. This study evaluates 32 machine-learning model configurations using 828 incident records from the G2504 Hangzhou Ring Freeway in China. Four fixed feature sets containing the top 6, 11, 16, and 21 predictors were defined from a Minimum Redundancy Maximum Relevance (MRMR) ranking. Each feature setting was evaluated using a 15% hold-out test set, seven repeated random splits, and sevenfold cross-validation within the training subset. All 32 machine-learning models were tuned within the training subset using sevenfold cross-validation. To provide a concise description of the tuning procedure, the selected Medium Gaussian support vector machine, which provided the most balanced overall performance, is reported as a representative example. Performance was assessed using the mean absolute error (MAE) and the mean absolute percentage error (MAPE). The Medium Gaussian support vector machine provided the most balanced overall performance, and prediction accuracy generally improved from 6 to 16 predictors, with little additional benefit from 21 predictors. Compared with classical linear regression, the Medium Gaussian support vector machine reduced MAE by 5.7%–9.7% and MAPE by 21.2%–31.9% across the four feature settings. A sensitivity comparison showed substantially higher errors when incidents longer than 60 minutes were retained, reinforcing that the main findings apply to routine incidents. Interpretation analyses identified Response Time and severity-related variables as important predictors, while residual errors remained larger for longer and more complex incidents. The model is intended primarily for post-arrival updating and clearance-support decisions for routine freeway incidents, rather than prediction at the initial alarm stage.
Related
- Serum proteins predict incident inflammatory bowel disease 14 years before diagnosis: a machine-learning and Cox modelling in the UK Biobank
- Integrating survival analysis and machine learning for modeling overtaking duration on rural two-lane highways
- Physics-guided explainable machine learning for predicting temporal scour evolution around hybrid spur dikes under unsteady flood hydrographs
- Predicting forage yield and water use efficiency using machine learning across diverse forage species under varying irrigation and fertilizer management in the Ethiopian highlands
- Predicting Lifespan-Extending Compounds Using Supervised Machine Learning
BibTeX
@article{ju2026predicting,
title = {{Predicting freeway incident duration using machine learning}},
author = {Yanni Ju and Jingqi Liu and Mostafa K. Ardakani and Wanqiu Li and Gen Li},
journal = {PLOS One},
year = {2026},
month = sep,
doi = {10.1371/journal.pone.0359353},
url = {https://doi.org/10.1371/journal.pone.0359353},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Predicting freeway incident duration using machine learning
AU - Yanni Ju
AU - Jingqi Liu
AU - Mostafa K. Ardakani
AU - Wanqiu Li
AU - Gen Li
PY - 2026
DA - 2026/09/30
JO - PLOS One
DO - 10.1371/journal.pone.0359353
UR - https://doi.org/10.1371/journal.pone.0359353
AB - Accurate prediction of freeway incident duration is important for traffic management and emergency resource allocation. This study evaluates 32 machine-learning model configurations using 828 incident records from the G2504 Hangzhou Ring Freeway in China. Four fixed feature sets containing the top 6, 11, 16, and 21 predictors were defined from a Minimum Redundancy Maximum Relevance (MRMR) ranking. Each feature setting was evaluated using a 15% hold-out test set, seven repeated random splits, and sevenfold cross-validation within the training subset. All 32 machine-learning models were tuned within the training subset using sevenfold cross-validation. To provide a concise description of the tuning procedure, the selected Medium Gaussian support vector machine, which provided the most balanced overall performance, is reported as a representative example. Performance was assessed using the mean absolute error (MAE) and the mean absolute percentage error (MAPE). The Medium Gaussian support vector machine provided the most balanced overall performance, and prediction accuracy generally improved from 6 to 16 predictors, with little additional benefit from 21 predictors. Compared with classical linear regression, the Medium Gaussian support vector machine reduced MAE by 5.7%–9.7% and MAPE by 21.2%–31.9% across the four feature settings. A sensitivity comparison showed substantially higher errors when incidents longer than 60 minutes were retained, reinforcing that the main findings apply to routine incidents. Interpretation analyses identified Response Time and severity-related variables as important predictors, while residual errors remained larger for longer and more complex incidents. The model is intended primarily for post-arrival updating and clearance-support decisions for routine freeway incidents, rather than prediction at the initial alarm stage.
SN - 1932-6203
ER -