摘要
由于车辆使用频率的日益增长和人民对于驾驶安全重视程度的不断提高,驾驶行为分析已经成为提升驾驶安全的重要一环。针对驾驶行为的耦合性、随机性和复杂性,文章通过C4.5算法度量驾驶行为的属性特征建立决策树,通过建立多个决策树,构成随机森林模型来提高驾驶行为分析的准确性。在计算机上使用WEKA软件完成算法的实现,最终验证了算法能够对驾驶行为实现准确的分析预测。
As a result of the increasing frequency of vehicle use and the increasing importance that people attach to driving safety,driving behavior analysis has become an important part of improving driving safety.According to the coupling,randomness and complexity of driving behavior,the C4.5 algorithm is used to measure the attribute of driving behavior to build a decision tree,and multiple decision trees were constructed to form a random forest model to improve the accuracy of driving behavior analysis.The realization of the algorithm using WEKA software on the computer finally proved that the algorithm can achieve accurate analysis and prediction of driving behavior.
作者
黄亮
董智浩
张锐明
Huang Liang;Dong Zhihao;Zhang Ruiming(Wuhan University of Technology,Wuhan 430070,China;Guangdong Tailuosi Power System Co.,Ltd.,Foshan 528200,China)
出处
《无线互联科技》
2018年第7期72-76,共5页
Wireless Internet Technology
基金
国家自然科学基金
项目编号:51477125
作者简介
黄亮(1979—),男,湖北武汉人,副教授,博士;研究方向:电力电子,物联网。;通信作者:董智浩(1993—),男,福建福州人,硕士;研究方向:物联网。
