Forecasting Fatal Traffic Accidents During Early Covid-19 And Increased Enforcement Activities: A Case Study from Hatay Province, Türkiye Using Classical Time Series and Ensemble Machine Learning Models

Kevser Unsalan; Mehmet Fatih Can; Omer Faruk Cansiz

doi:10.7250/bjrbe.2025-20.664

Forecasting Fatal Traffic Accidents During Early Covid-19 And Increased Enforcement Activities: A Case Study from Hatay Province, Türkiye Using Classical Time Series and Ensemble Machine Learning Models

Authors

Kevser Unsalan Department of Civil Engineering, Iskenderun Technical University, Hatay, Türkiye https://orcid.org/0000-0002-9163-4855
Mehmet Fatih Can Department of Water Resources Management, Iskenderun Technical University, Hatay, Türkiye https://orcid.org/0000-0002-3866-2419
Omer Faruk Cansiz Department of Civil Engineering, Iskenderun Technical University, Hatay, Türkiye https://orcid.org/0000-0001-6857-2513

DOI:

https://doi.org/10.7250/bjrbe.2025-20.664

Keywords:

ensemble machine learning, fatal traffic accidents, time series forecasting, traffic enforcement

Abstract

This study investigates the temporal dynamics and predictive modelling of fatal traffic accidents in Hatay province, Türkiye, using both classical time series approaches (ARIMA, SARIMA, Holt–Winters) and machine learning techniques (Random Forest, Gradient Boosting). Monthly accident data from 2017–2021 were analysed through seasonal decomposition, stationarity testing, and comparative model evaluation. Results revealed a distinct seasonal pattern, with accident counts peaking during summer months and declining in winter, and a long-term trend showing a notable reduction in fatalities after 2017. Among the tested models, the Enhanced Gradient Boosting approach demonstrated the highest predictive accuracy (R² = 0.97, RMSE = 1.59), outperforming both classical time series and other ensemble methods. Forecast results for 2021 indicated seasonal peaks in June and August, corresponding to increased traffic density during the holiday period. The COVID-19 pandemic was associated with a marked short-term reduction in fatalities, though the effect appeared to diminish post-lockdown. These findings highlight the value of integrating advanced ensemble learning methods into traffic safety forecasting and underscore the importance of seasonally targeted interventions.

Supporting Agencies

General Directorate of Security of the Ministry of Internal Affairs of the Republic of Turkey

References

Ahmed, N., Mahmood, A. N., & Hu, J. (2020). A survey of network anomaly detection techniques. Journal of Network and Computer Applications, 60, 19–31. https://doi.org/10.1016/j.jnca.2015.11.016 DOI: https://doi.org/10.1016/j.jnca.2015.11.016

Basher, S. A., & Sadorsky, P. (2022). Forecasting Bitcoin price direction with random forests: How important are interest rates, inflation, and market volatility? Machine Learning with Applications, 9, Article 100355. https://doi.org/10.1016/j.mlwa.2022.100355 DOI: https://doi.org/10.1016/j.mlwa.2022.100355

Başeğmez, M., & Aydın C. C. (2022). Türkiye’de COVID-19 sürecinde alınan önlemler çerçevesinde okul bahçe ve sınıflarının CBS ile değerlendirilmesi. Geomatik, 7(3), 209–219. https://doi.org/10.29128/geomatik.971403 DOI: https://doi.org/10.29128/geomatik.971403

Box, G. E. P., Jenkins, G. M., Reinsel, G. C., & Ljung, G. M. (2015). Time series analysis: Forecasting and control (5th ed.). Wiley.

Chen, T., & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 785–794. https://doi.org/10.1145/2939672.2939785 DOI: https://doi.org/10.1145/2939672.2939785

Çiftçi, S., & Batur Sir, G. D. (2023). An application for forecasting the number of applications to the emergency department with time series analysis and machine learning methods. Pamukkale University Journal of Engineering Sciences, 29(7), 667–679. https://doi.org/10.5505/pajes.2022.18488 DOI: https://doi.org/10.5505/pajes.2022.18488

Ergin, H., & Güler, H. (2025). Türkiye’de trafik para cezalarının trafik kazaları üzerindeki etkisi. Omer Halisdemir Universitesi Iktisadi ve Idari Bilimler Fakültesi Dergisi, 18(2), 706–720. https://doi.org/10.25287/ohuiibf.1602514 DOI: https://doi.org/10.25287/ohuiibf.1602514

Friedman, J. H. (2001). Greedy function approximation: A gradient boosting machine. Annals of Statistics, 29(5), 1189–1232. https://doi.org/10.1214/aos/1013203451 DOI: https://doi.org/10.1214/aos/1013203451

Getahun, K. A. (2021). Time series modelling of road traffic accidents in Amhara Region. Journal of Big Data, 8(1), Article 102. https://doi.org/10.1186/s40537-021-00493-z DOI: https://doi.org/10.1186/s40537-021-00493-z

Hyndman, R. J., & Athanasopoulos, G. (2021). Forecasting: Principles and practice (3rd ed.). OTexts.

Hyndman, R. J., & Khandakar, Y. (2008). Automatic time series forecasting: the forecast package for R. Journal of Statistical Software, 27(3), 1–22. https://doi.org/10.18637/jss.v027.i03 DOI: https://doi.org/10.18637/jss.v027.i03

Highway Traffic Accident Statistic. https://data.tuik.gov.tr/Bulten/Index?p=Karayolu-Trafik-Kaza-Istatistikleri-2024-54056

Ke, G., Meng, Q., Finley, T., Wang, T., Chen, W., Ma, W., Ye, Q., & Liu, T. Y. (2017). Light GBM: A highly efficient gradient boosting decision tree. 31st Conference on Neural Information Processing Systems (NIPS 2017), Long Beach, CA, USA. https://proceedings.neurips.cc/paper_files/paper/2017/file/6449f44a102fde848669bdd9eb6b76fa-Paper.pdf

Li, L., Dai, S., Cao, Z., Hong, J., Jiang, S., & Yang, K. (2020). Using improved gradient-boosted decision tree algorithm based on Kalman filter (GBDT-KF) in time series prediction. Journal of Supercomputing, 76(9), 6887–6900. https://doi.org/10.1007/s11227-019-03130-y DOI: https://doi.org/10.1007/s11227-019-03130-y

Li, H., Liu, S., & Wu, D. (2022). A hybrid deep learning model for traffic accident prediction using spatiotemporal features. Applied Soft Computing, 123, Article 108938.

Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Muller, A., Nothman, J., Louppe, G., Prettenhofer, P., Weiss, R., Dubourg, V., Passos, A., Cournapeau, A., Brucher, M., Perrot, M., & Duchesnay, É. (2011). Scikit-learn: Machine learning in Python. Journal of Machine Learning Research, 12, 2825–2830. https://jmlr.org/papers/volume12/pedregosa11a/pedregosa11a.pdf

Perron, P. (2006). Dealing with structural breaks. In T. C. Mills & K. Patterson (Eds.), Palgrave handbook of econometrics (Vol. 1, pp. 278–352). Palgrave Macmillan.

Popescu, T. D. (2020). Time series analysis for assessing and forecasting of road traffic accidents – Case Studies. WSEAS Transactions on Mathematics, 19, 177–186. https://doi.org/10.37394/23206.2020.19.17 DOI: https://doi.org/10.37394/23206.2020.19.17

Rodríguez, J. M., Peñaloza, R. E., & Moreno Montoya, J. (2015). Road traffic injury trends in the city of Valledupar, Colombia: A time series study from 2008 to 2012. PLoS ONE, 10(12), Article e0144002. https://doi.org/10.1371/journal.pone.0144002 DOI: https://doi.org/10.1371/journal.pone.0144002

Seabold, S., & Perktold, J. (2010). Econometric and statistical modeling with Python. Proc. of the 9th Python in Science Conf. (SCIPY 2010), 57–61. https://www.researchgate.net/publication/264891066_Statsmodels_Econometric_and_Statistical_Modeling_with_Python DOI: https://doi.org/10.25080/Majora-92bf1922-011

Sheiso, D. S., Ware, A. N., & Woyeso, N. M. (2024). Time series modelling of road traffic accidents in West Arsi, Ethiopia. Journal of Advance Research in Mathematics and Statistics, 11(1), 48–65. https://www.researchgate.net/publication/382929799_TIME_SERIES_ MODELLING_OF_ROAD_TRAFFIC_ACCIDENTS_IN_WEST_ARSI_ETHIOPIA DOI: https://doi.org/10.61841/3t8atv54

Shilling, F., Waetjen, D., & Li, J. (2021). Dramatic reductions in traffic crashes, injuries and fatalities during COVID-19 stay-at-home period in California. Accident Analysis & Prevention, 153, Article 106027.

Sjösten, L. (2022). A comparative study of the KPSS and ADF tests in terms of size and power [Bachelor’s thesis in Statistics, Uppsala Universitat]. https://www.diva-portal.org/smash/get/diva2:1668033/FULLTEXT01.pdf

Sümer, N., & Kaygisiz, Ö. (2015). The interplay between enforcement, tickets issued, and traffic safety in Turkey: Analyses of 2008-2012. Turkish Journal of Public Health, 13(3), 193-205. https://doi.org/10.20518/thsd.11797 DOI: https://doi.org/10.20518/thsd.11797

Yan, M., & Shen, Y. (2022). Traffic accident severity prediction based on random forest. Sustainability, 14(3), Article 1729. https://doi.org/10.3390/su14031729 DOI: https://doi.org/10.3390/su14031729

Yousefzadeh-Chabok, S., Ranjbar-Taklimie, F., Malekpouri, R., & Razzaghi, A. (2016). A time series model for assessing the trend and forecasting the road traffic accident mortality. Archives of Trauma Research, 5(3), 1–6. https://doi.org/10.5812/atr.36570 DOI: https://doi.org/10.5812/atr.36570

Zhao, Y., Lin, T., & Xu, M. (2022). Impact of COVID-19 on traffic accidents in urban China: A time series analysis. Journal of Transport & Health, 25, Article 101335.