Prediction models for understanding traffic safety risk on rural roads in Serbia

Abstract

This paper presents a traffic accident prediction model developed for IB-category rural roads in the Republic of Serbia, utilizing the largest dataset analyzed in the country to date. The model employs interpretable machine learning techniques to assess the impact of road geometry, traffic exposure, and quality on accident frequency. The model exhibits robust predictive accuracy with a R² of 79%, especially in pinpointing high-risk regions linked to increased traffic density and geometric limitations. SHAP-cluster analysis was used to interpret variable contributions and reveal hidden patterns in cluster-specific risk profiles, which can be used to define and prioritize prevention measures. Future study will expand this framework by extending it to various road types, including freeways, and by investigating additional elements, such as user behavior and weather conditions, that may improve forecast accuracy. The results provide a practical tool for road authorities and policymakers aiming to improve traffic safety on the rural road network.

Keywords

Traffic accident prediction / Rural roads / Road safety modeling / Machine learning

References

[1] Ahmed S, Hossain MA, Ray SK, Bhuiyan MMI, Sabuj SR. A study on road accident prediction and contributing factors using explainable machine learning models: Analysis and performance. Transportation Research Interdisciplinary Perspectives. 2023, 19(April): 100814.

[2] Alagarsamy, S., Malathi, M., Manonmani, M., Sanathani, T., & Kumar, A. (2021). Prediction of road accidents using machine learning technique. 5th International Conference on Electronics, Communication and Aerospace Technology (ICECA) (pp. 1695–1701). https://doi.org/10.1109/ICECA52323.2021.9675852

[3] Almanie T. Quantitative study of traffic accident prediction models: A case study of Virginia accidents. The International Journal of Advanced Networking and Applications. 2023, 14055582-5589.

[4] Ameksa, M., Mousannif, H., Al Moatassime, H., & Elamrani Abou Elassad, Z. (2022). Crash prediction using ensemble methods. BML 2021 - International Conference on Big Data, Modelling and Machine Learning (pp. 211–215). https://doi.org/10.5220/0010731200003101

[5] Anastasopoulos PC, Mannering FL. A note on modeling vehicle accident frequencies with random-parameters count models. Accident Analysis and Prevention. 2009, 41(1): 153-159.

[6] Ariannezhad A, Karimpour A, Qin X, Wu YJ, Salmani Y. No title. Journal of Transportation Engineering, Part a: Systems. 2020, 147(3): 16.

[7] Asadi M, Ulak MB, Geurs KT, Weijermars W, Schepers P. A comprehensive analysis of the relationships between the built environment and traffic safety in the Dutch urban areas. Accident Analysis and Prevention. 2022, 172November 2021106683.

[8] Astarita V, Haghshenas SS, Guido G, Vitale A. Developing new hybrid grey wolf optimization-based artificial neural network for predicting road crash severity. Transportation Engineering. 2023, 12September 2022100164.

[9] Augustine, T., & Shukla, S. (2022). Road accident prediction using machine learning approaches. 2nd International Conference on Advance Computing and Innovative Technologies in Engineering (ICACITE) (pp. 808–811). https://doi.org/1PNDV7scbonS1t5xCBm2ncuVNECtXhmJs

[10] Awasthi D, Parti R, Mahajan K. Effect of spatial relationship between curves on crash severity at horizontal curves in a mountainous terrain. Accident Analysis and Prevention. 2024, 206October 2024107714.

[11] Basso F, Basso LJ, Pezoa R. The importance of flow composition in real-time crash prediction. Accident Analysis and Prevention. 2020, 137January105436.

[12] Behboudi, N., Moosavi, S., & Ramnath, R. (2024). Recent advances in traffic accident analysis and prediction: a comprehensive review of machine learning techniques. ArXiv, 2406.13968, 26. https://doi.org/10.48550/arXiv.2406.13968

[13] Bhattarai N, Lin C, Zhang Y, Liu H. In-depth investigation of contributing factors of fatal/severe-injury crashes at highway merging areas using machine learning classification methods. Journal of Traffic and Transportation Engineering (English Edition). 2025, 122434-446.

[14] Borsos A. Accident prediction models for Hungarian two-lane rural first-class main roads. Acta Technica Jaurinensis. 2014, 7(3): 280-293.

[15] Cadar RD, Boitor MR, Dumitrescu M. Effects of traffic volumes on accidents: The case study of Romania’s national roads. Geographia Technica. 2017, 12(2): 20-29. 10.21163/GT

[16] Charm T, Wang H, Zuniga-Garcia N, Ahmed M, Kockelman KM. Predicting crash occurrence at intersections in Texas: An opportunity for machine learning. Transportation Planning and Technology. 2024, 47(8): 1184-1204.

[17] Chen C, Xie Y. Modeling the effects of AADT on predicting multiple-vehicle crashes at urban and suburban signalized intersections. Accident Analysis and Prevention. 2016, 91: 72-83.

[18] Choi JG, Kong CW, Kim G, Lim S. Car crash detection using ensemble deep learning and multimodal data from dashboard cameras. Expert Systems with Applications. 2021, 183: 115400.

[19] Choudhary A, Garg RD, Jain SS. Estimating impact of pavement surface condition and geometrics design on two-wheeler run-off road crashes on horizontal curves. Iatss Research. 2024, 48(1): 108-119.

[20] Čubranić-Dobrodolac M, Lipovac K, Čičević S, Antić B. A model for traffic accidents prediction based on driver personality traits assessment. Promet-Traffic and Transportation. 2017, 296631-642.

[21] Das S, Sun X, Sun M. Rule-based safety prediction models for rural two-lane run-off-road crashes. International Journal of Transportation Science and Technology. 2021, 10(3): 235-244.

[22] Dereli MA, Erdogan S. A new model for determining the traffic accident black spots using GIS-aided spatial statistical methods. Transportation Research Part a: Policy and Practice. 2017, 103: 106-117.

[23] Deretić N, Stanimirović D, Al Awadh M, Vujanović N, Djukić A. SARIMA modelling approach for forecasting of traffic accidents. Sustainability (Switzerland). 2022, 14(8): 4403.

[24] Dong S, Khattak A, Ullah I, Zhou J, Hussain A. Predicting and analyzing road traffic injury severity using boosting-based ensemble learning models with SHAPley additive explanations. International Journal of Environmental Research and Public Health. 2022, 19523.

[25] Dutta N, Fontaine MD. Improving freeway segment crash prediction models by including disaggregate speed data from different sources. Accident Analysis and Prevention. 2019, 132July105253.

[26] Elghriany A, Yi P, Liu P, Yu Q. Investigation of the effect of pavement roughness on crash rates for rigid pavement. Journal of Transportation Safety & Security. 2016, 82164-176.

[27] Elyassami, S., Hamid, Y., & Habuza, T. (2021). Road crashes analysis and prediction using gradient boosted and random forest trees. 2020 6th IEEE Congress on Information Science and Technology (CiSt) (p. 7). https://doi.org/10.1109/CiSt49399.2021.9357298

[28] Gatarić D, Ruškić N, Aleksić B, Đurić T, Pezo L, Lončar B, Pezo M. Predicting road traffic accidents—artificial neural network approach. Algorithms. 2023.

[29] Geyik, B., & Kara, M. (2020). Severity prediction with machine learning methods. HORA 2020 - 2nd International Congress on Human-Computer Interaction, Optimization and Robotic Applications, Proceedings (p. 7). https://doi.org/10.1109/HORA49412.2020.9152601

[30] Glavić D, Mladenović M, Stevanovic A, Tubić V, Milenković M, Vidas M. Contribution to accident prediction models development for rural two-lane roads in Serbia mark as interesting comment. Promet-Traffic & Transportation. 2016, 28(4): 415-424.

[31] Goh K, Currie G, Sarvi M, Logan D. Factors affecting the probability of bus drivers being at-fault in bus-involved accidents. Accident Analysis and Prevention. 2014, 66: 20-26.

[32] He M, Meng G, Wu X, Han X, Fan J. Road traffic accident prediction based on multi-source data – a systematic review. Promet-Traffic and Transportation. 2025, 37(2): 499-522.

[33] Hou, X. (2023). Urban traffic accident factor analysis, risk prediction and early warning based on big data technology. ICCAI ‘23: Proceedings of the 2023 9th International Conference on Computing and Artificial Intelligence (pp. 691–696). https://doi.org/10.1145/3594315.3594392

[34] Hsu CK. Reconsidering seasonality, weather, and road safety in non-temperate areas: The case of Kaohsiung, Taiwan. Travel Behaviour and Society. 2024, 34(July 2023): 100710.

[35] Hussain SF, Ashraf MM. A novel one-vs-rest consensus learning method for crash severity prediction. Expert Systems with Applications. 2023, 228May120443.

[36] Hyodo S, Hasegawa K. Factors affecting analysis of the severity of accidents in cold and snowy areas using the ordered probit model. Asian Transport Studies. 2021, 7February100035.

[37] Karacasu M, Ergül B, Yavuz AA. Estimating the causes of traffic accidents using logistic regression and discriminant analysis. International Journal of Injury Control and Safety Promotion. 2014, 214305-313.

[38] Khattak AJ, Liu J, Wali B, Li X, Ng MW. Modeling traffic incident duration using quantile regression. Transportation Research Record. 2016, 2554(2554): 139-148.

[39] Labib, F., Rifat, A. S., Hossain, M., Das, A. K., & Nawrine, F. (2019). 2019 7th International Conference on Smart Computing and Communications, ICSCC 2019. 2019 7th International Conference on Smart Computing and Communications, ICSCC 2019 (pp. 1–5).

[40] Li Y, Liu C, Ding L. Impact of pavement conditions on crash severity. Accident Analysis and Prevention. 2013, 59: 399-406.

[41] Li G, Wu Y, Bai Y, Zhang W. ReMAHA–CatBoost: Addressing imbalanced data in traffic accident prediction tasks. Applied Sciences (Switzerland). 2023, 13(24): 22.

[42] Li J, Guo F, Zhou Y, Yang W, Ni D. Predicting the severity of traffic accidents on mountain freeways with dynamic traffic and weather data. Transportation Safety and Environment. 2023, 5414.

[43] Li T, Klavins J, Xu T, Zafri NM, Stern R. Identifying built environment factors influencing driver yielding behavior at unsignalized intersections: A naturalistic open-source dataset collected in Minnesota. Journal of Safety Research. 2025, 92April 2024331-345.

[44] Ma Z, Mei G, Cuomo S. An analytic framework using deep learning for prediction of traffic accident injury severity based on contributing factors. Accident Analysis and Prevention. 2021, 160(9): 106322.

[45] Mićić S, Vujadinović R, Amidžić G, Damjanović M, Matović B. Accident frequency prediction model for flat rural roads in Serbia. Sustainability (Switzerland). 2022, 14137704.

[46] Mujalli RO, Al-Masaeid H, Alamoush S. Modeling traffic crashes on rural and suburban highways using ensemble machine learning methods. KSCE Journal of Civil Engineering. 2023, 272814-825.

[47] Nasri M, Aghabayk K. Assessing risk factors associated with urban transit bus involved accident severity: A case study of a Middle East country. International Journal of Crashworthiness. 2021, 26(4): 413-423.

[48] Parra, C., Ponce, C., & Rodrigo, S. F. (2020). Evaluating the performance of explainable machine learning models in traffic accidents prediction in California. Proceedings - International Conference of the Chilean Computer Science Society, SCCC, 2020-Novem, 2–9. https://doi.org/10.1109/SCCC51225.2020.9281196

[49] Pavlou D, Christodoulou G, Yannis G. The impact of weather conditions and driver characteristics on road safety on rural roads. Transportation Research Procedia. 2023, 72: 4081-4088.

[50] Pitarque A, Guillen M. Interpolation of quantile regression to estimate driver’s risk of traffic accident based on excess speed. Risks. 2022, 10(1): 13.

[51] Pljakić M, Jovanović D, Matović B, Mićić S. Macro-level accident modeling in Novi Sad: A spatial regression approach. Accident Analysis and Prevention. 2019, 132July105259.

[52] Pljakić, M., Stanojević, P., & Jovanović, D. (2021). Prediction of accident using time series methods and models in Serbia. VIII International Symposium “New Horizons 2021,” (p. 11). https://www.researchgate.net/publication/361746186

[53] Pljakić, M., Dragan, J., Predrag, S., Svetlana, B., & Boško, M. (2022). Forecasting fatalities and injuries of accidents using the arima. 17. International Conference “Road Safety in Local Community,” July, 10. https://www.researchgate.net/publication/361746255. Forecasting

[54] Pljakić, M., Lindov, O., Petrović, A., Stanojević, P., & Arsić, N. (2024). Application of machine learning for prediction of road accidents based on indicators: A random forest approach. International Conference on Advances in Traffic and Communication Technologies (ATCT) (pp. 70–76). www.atct.ba

[55] Rahman MA, Sun X, Das S, Khanal S. Exploring the influential factors of roadway departure crashes on rural two-lane highways with logit model and association rules mining. International Journal of Transportation Science and Technology. 2021, 102167-183.

[56] Ranković, V., Đonić, A., & Geroski, T. (2024). Road traffic accidents prediction using machine learning methods. 10th International Congress - Motor Vehicles & Motors 2024 (pp. 409–415). https://doi.org/978-86-6335-120-2

[57] Road Traffic Safety Agency. (2023). Statistical report on traffic safety in Serbia. https://doi.org/2334-9395

[58] Rúa E, Arias P, Saavedra Á, Martínez-Sánchez J. Combination of macroscopic and microscopic crash prediction models with multiple modeling approaches: A highway case study. Expert Systems with Applications. 2024, 258(February): 125158.

[59] ShaffieeHaghshenas S, Guido G, ShaffieeHaghshenas S, Astarita V. Predicting the level of road crash severity: A comparative analysis of logit model and machine learning models. Transportation Engineering. 2025, 20November 2024100323.

[60] Shaik ME, Islam MM, Hossain QS. A review on neural network techniques for the prediction of road traffic accident severity. Asian Transport Studies. 2021, 7November100040.

[61] Silva PB, Andrade M, Ferreira S. Influence of segment length on the fitness of multivariate crash prediction models applied to a Brazilian multilane highway. Iatss Research. 2021, 454493-502.

[62] Sl. glasnik RS br. 50/11. (2011). Pravilnik o uslovima koje sa aspekta bezbednosti saobraćaja moraju da ispunjavaju putni objekti i drugi elementi javnog puta. http://demo.paragraf.rs/demo/combined/Old/t/t2011_07/t07_0139.htm

[63] Službeni glasnik RS br. 30/18. (2018). Traffic safety strategy of the Republic of Serbia for the period 2023 to 2030. https://www.abs.gov.rs/static/uploads/15117_strategija-bezbednosti-saobracaja-2023-2030.pdf

[64] Sun J, Sun J. A dynamic Bayesian network model for real-time crash prediction using traffic speed conditions data. Transportation Research Part c: Emerging Technologies. 2015, 54: 176-186.

[65] Tamim Kashifi M, Ahmad I. Efficient histogram-based gradient boosting approach for accident severity prediction with multisource data. Transportation Research Record: Journal of the Transportation Research Board. 2022, 2676(6): 236-258.

[66] Venkat A, Vijey GK, Susan Thomas I. Machine learning based analysis for road accident prediction. International Journal of Emerging Technology and Innovative Engineering. 2020, 6(02): 2394-6598https://ssrn.com/abstract=3550462

[67] Vorko-Jović A, Jović F. Macro model prediction of elderly people’s injury and death in road traffic accidents in Croatia. Accident Analysis and Prevention. 1992, 24(6): 667-672.

[68] Wang L, Abdel-Aty M, Shi Q, Park J. Real-time crash prediction for expressway weaving segments. Transportation Research Part c: Emerging Technologies. 2015, 61: 1-10.

[69] Xiao D, Ding H, Sze NN, Zheng N. Investigating built environment and traffic flow impact on crash frequency in urban road networks. Accident Analysis and Prevention. 2024, 201(April): 107561.

[70] Xie Y, Zhao K, Huynh N. Analysis of driver injury severity in rural single-vehicle crashes. Accident Analysis and Prevention. 2012, 47: 36-44.

[71] Xu X, Duan L. Predicting crash rate using logistic quantile regression with bounded outcomes. IEEE Access. 2017, 5January27036-27042.

[72] Yocum RL, Gayah VV. County-level crash prediction models for Pennsylvania accounting for income characteristics. Transportation Research Interdisciplinary Perspectives. 2022, 13: 100562.

[73] Zeng L, Hu Z, Sayed T. Traffic conflict prediction at signal cycle level using Bayesian optimized machine learning approaches. Transportation Research Record. 2022, 26775183-195.

[74] Zhang W, Liu T, Yi J. Exploring the spatiotemporal characteristics and causes of rear-end collisions on urban roadways. Sustainability (Switzerland). 2022, 141823.

[75] Zhang Z, Nie Q, Liu J, Hainen A, Islam N, Yang C. Machine learning based real-time prediction of freeway crash risk using crowdsourced probe vehicle data. Journal of Intelligent Transportation Systems. 2022, 28: 84-102.

[76] Zhang C, Yang X, Huang J, Xiao Z. Uncovering spatial patterns of environmental influence on the paces of active leisure travel. Cities. 2025, 162July105971.

[77] Zhu L, Lu L, Zhang W, Zhao Y, Song M. Analysis of accident severity for curved roadways based on bayesian networks. Sustainability (Basel). 2019, 11817.

[78] Zou X, Sun L, Lan T, Fan C, Liu S, Zhao H, Qiu J. The effects of weather factors on road traffic casualties: Analysis on provincial panel data of China from 2006 to 2021. Heliyon. 2024, 1017e36788.

RIGHTS & PERMISSIONS

The Author(s)