Abstract
In order to further explore the occurrence mechanism of road traffic accidents; this paper takes into account the differences in road types. Based on the open source traffic accident data set of the British Ministry of Transport, the traffic accident risk factors and their importance of the four road types with the highest traffic accident rates are studied. Firstly, based on data visualization, the distribution of data is preliminarily explored from the spatial and temporal dimensions. Secondly, aiming at the shortcomings of the integrated algorithm with more hyperparameters, which lacks accurate and efficient parameter adjustment methods, a Bayesian optimization based random forest algorithm (BO-RF) implemented by the Optuna framework is proposed to construct the risk factor importance level screening model of four road types. In addition, a control experiment with the random forest algorithm based on random search optimization (RS-RF) was carried out. The results show that the BO-RF algorithm model based on the Optuna hyperparameter optimization framework has shorter tuning time and higher accuracy. Finally, based on the visualization results of road traffic accident risk factor level obtained by the model constructed by BO-RF algorithm, a scheme suitable for reducing accidents under this road condition is proposed, which provides a reference for preventing road traffic accidents and optimizing relevant safety management regulations.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Wu, G.F., Liu, F.J., Dong, G.L.: Analysis of the influencing factors of road environment in road traffic accidents. In: Changsha: 2020 4th Annual International Conference on Data Science and Business Analytics (ICDSBA), pp. 83–85. IEEE
Pawlowski, W., Lasota, D., Goniewicz, M., et al.: The effect of ethyl alcohol upon pedestrian trauma sustained in traffic crashes. Int. J. Environ. Res. Public Health 16(8), 1471 (2019)
Zhan, Q.J.: Based Railway Accident Casual Factor Modeling and Hybrid Learning Approach. Beijing Jiaotong University, Beijing (2017)
Kuniyuki, H., Zainuddin, A.F.: A study on method of extracting high-risk accident causes in Nagano using Bayesian network analysis. Int. J. Autom. Engin. 13(3), 139–146 (2022)
Su, F., Yuan, Q.: Analysis on association rules of urban road traffic accidents. J. Wuhan Univ. Technol. (Inf. Manag. Engin.) 42(04), 313–318+331
Wang, X., Han, T.: Transformer fault diagnosis based on bayesian optimized random forest. Electr. Measure. Instrum. 58(06), 167–173 (2021)
Yang, C., Liu, L.L., Zhang, Y.L., et al.: Landslide susceptibility assessment based on Bayesian optimized machine learning hyperparameters. Geol. Technol. Bull. 41(02), 228–238 (2022)
Ma, H., Ren, M.X., Gao, K.: Gas emission prediction model based on random search optimization XGBoost. J. Safety Technol. 18(05), 129–134 (2022)
Shahriari, B., Swersky, K., Wang, Z., Adams, R.P., Freitas, N.D.: Taking the human out of the loop: A review of Bayesian optimization. Proc. IEEE 104(1), 148–175 (2016)
Cui, J.X., Yang, B.: Overview of Bayesian optimization methods and applications. J. Softw. 29(10), 3068–3090 (2018)
UK Road Safety Dataset: https://www.data.gov.uk/dataset/cb7ae6f0-4be6-4935-9277-47e5ce24a11f/road-safety-data
Wang, M., Liu, D.: Cause analysis of railway traffic accidents based on random forest. J. Phys.: Confer. Ser. IOP Publishing 1910(1), 012017
Li, Y.S., Zhang, X., Wang, W.J., et al.: Analysis on influencing factors of accident injury degree of electric bicycle riders based on random forest. J. Transp. Syst. Eng. Inf. Technol. 21(01), 196–200 (2021)
Zhang, Y.: Research on Hyperparameter Optimization of Deep Learning Model. Capital University of Economics and Business, Beijing (2020)
De Ath, G., Everson, R.M., Rahat, A.A.M., et al.: Greed is good: exploration and exploitation trade-offs in Bayesian optimisation. ACM Trans. Evolut. Learn. Optim. 1(1), 1–22 (2021)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2024 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Jie, W., Zhenfei, Z., Liuzhu, Q. (2024). Study on Grade Discrimination Method of Traffic Accident Risk Factors Considering Road Type. In: Arai, K. (eds) Intelligent Systems and Applications. IntelliSys 2023. Lecture Notes in Networks and Systems, vol 824. Springer, Cham. https://doi.org/10.1007/978-3-031-47715-7_29
Download citation
DOI: https://doi.org/10.1007/978-3-031-47715-7_29
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-47714-0
Online ISBN: 978-3-031-47715-7
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)