科技创新与应用2024，Vol.14Issue(36)：17-22,6.DOI:10.19981/j.CN23-1581/G3.2024.36.004
基于H2O自动化机器学习的电动自行车交通违法行为影响因素分析
申远 1戴帅 1赵琳娜 1杨钧剑 1侯志贤1
作者信息
摘要
Abstract
As the number of electric bicycles increases year by year,the proportion of electric bicycle traffic accident deaths in the total number of traffic accident deaths continues to increase.In this context,exploring the factors influencing electric bicycle illegal behaviors is important for reducing accident rates and accident severity.significance.This study used data on illegal behaviors of electric bicycles in Xiamen Island to analyze the influencing factors of illegal behaviors of electric bicycles based on the H2O automated machine learning(AutoML)algorithm,and compared it with the Random Forest(RF)algorithm.The results show that H2O automated machine learning has better prediction accuracy and efficiency.In addition,built environment variables were also introduced into the study.The research results show that built environments such as small shops and residential areas have a positive impact on illegal activities,and small shops dominate among many influencing factors.关键词
交通安全/自动化机器学习/电动自行车违法/建成环境/随机森林Key words
traffic safety/automated machine learning/illegal electric bicycles/built environment/Random Forest(RF)分类
交通工程引用本文复制引用
申远,戴帅,赵琳娜,杨钧剑,侯志贤..基于H2O自动化机器学习的电动自行车交通违法行为影响因素分析[J].科技创新与应用,2024,14(36):17-22,6.
