Archive

Published: Građevinar 77 (2025) 12
Paper type: Original scientific paper
Download article (Croatian): PDF
Download article (English): PDF

Predicting unsafe road sections using machine learning

Riste Ristov, Slobodan Ognjenovic, Zlatko Zafirovski

Abstract

This paper presents an ML methodology to predict hazardous road segments, using the weighted accident index (Wi). The analysis covers 161 road segments in North Macedonia (~1,300 km)—with 23+1 variables categorized into Road, Traffic, Environmental, and Accident data. Feature influence is evaluated using six models with an 80/20 training/testing split. Weighted SHAP is applied to obtain a single variable ranking; XGBoost with 15 inputs is the final predictor. The model achieves a validated performance (R² = 0.65), while operational prioritization yields AUROC = 0.69 at Wi ≥ 10.13, enabling timely identification of hazardous segments and interventions by relevant authorities.

Keywords
road safety, machine learning, prediction, SHAP, weighted accident index, traffic analysis

HOW TO CITE THIS ARTICLE:

Ristov, R., Ognjenovic, S., Zafirovski, Z.: Predicting unsafe road sections using machine learning, GRAĐEVINAR, 77 (2025) 12, pp. 1187-1199, doi: https://doi.org/10.14256/JCE.4319.2025

OR:

Ristov, R., Ognjenovic, S., Zafirovski, Z. (2025). Predicting unsafe road sections using machine learning, GRAĐEVINAR, 77 (12), 1187-1199, doi: https://doi.org/10.14256/JCE.4319.2025

LICENCE:

Creative Commons License
This paper is licensed under a Creative Commons Attribution 4.0 International License.
Authors:
Ristov WEB
Riste Ristov
University of St. Cyril and Methodius, N. Macedonia
Faculty of Civil Engineering
Ognjenovic WEB
Slobodan Ognjenovic
University of St. Cyril and Methodius, N. Macedonia
Faculty of Civil Engineering
Zafirovski WEB
Zlatko Zafirovski
University of St. Cyril and Methodius, N. Macedonia
Faculty of Civil Engineering