摘要
为研究电动自行车交通事故中影响电动自行车骑行者伤害程度的因素,利用一县级市 2018-2021 年涉电动自行车与机动车发生交通事故数据,选取22 类与交通事故相关的因素,运用随机森林算法对22 类变量进行筛选,采用SVM算法进行分类预测.结果表明:电动自行车交通事故致电动自行车骑行者死亡或重伤的重要因素依次为交通方式、驾龄、道路类型、头盔使用情况、违法行为、事故地点、路口路段类型、道路隔离、性别、交通控制方式.
Abstract
In order to study the factors that affect the degree of injury to electric bicycle riders in electric bicycle traffic accidents,22 types of factors related to traffic accidents were selected based on the data of electric bicycle and motor vehicle accidents in a county-level city from 2018 to 2021.Random forest algorithm was used to screen the 22 types of variables,and SVM algorithm was used for classification prediction.The results indicate that the important factors causing death or serious injury to electric bicycle riders in elec-tric bicycle traffic accidents are,in turn,traffic mode,driving experience,road type,helmet usage,illegal behavior,accident location,in-tersection section type,road isolation,gender,and traffic control method.
基金项目
河南警察学院2023年度科研项目(HNJY-2023-61)
河南省科技厅2023年度科技攻关项目(232102240019)
维普
万方数据