Analysis of Traffic Accidents Based on the Integration Model

作     者：Yanshun Ma Yi Shi Yihang Song Chenxiao Wu Yuanzhi Liu 

作者机构：The School of Civil Engineering and TransportationNortheast Forestry UniversityHarbin 150040Heilongjiang ProvinceChina 

出 版 物：《Journal of Electronic Research and Application》 (电子研究与应用)

年 卷 期：2024年第8卷第1期

页      面：51-59页

学科分类：0809[工学-电子科学与技术（可授工学、理学学位）] 08[工学] 

主　　题：Stacking integrated learning Data analysis Traffic safety 

摘      要：To enhance the safety of road traffic operations,this paper proposed a model based on stacking integrated learning utilizing American road traffic accident ***,the process involved data cleaning,transformation,and ***,various classification models were constructed,including logistic regression,k-nearest neighbors,gradient boosting,decision trees,AdaBoost,and extra trees *** metrics such as accuracy,precision,recall,F1 score,and Hamming loss were *** analysis,the passive-aggressive classifier model exhibited superior comprehensive indices compared to other *** on the model’s output results,an in-depth examination of the factors influencing traffic accidents was ***,measures and suggestions aimed at reducing the incidence of severe traffic accidents were *** findings served as a valuable reference for mitigating the occurrence of traffic accidents.