基于改进K-均值聚类算法的汽车用户行为分析方法研究

Research on behavior analysis of automobile users based on improved K-means clustering algorithm

汽车用户的驾驶行为和操作习惯等决定着驾驶是否存在风险,对于道路交通安全具有重要的意义,针对这种情况,提出一种用于预测汽车用户驾驶行为风险等级的模型。该模型为了提高模型的分类效率,在自组织映射神经网络算法中采用遗忘第二名的策略,然后结合自组织映射神经网络改进K-均值聚类分析方法,实现对于车辆驾驶人员的风险行为等级进行划分,通过聚类分析得到风险标签后,利用XGBoost算法实现对于用户风险行为的辨识。实验结果表明,改进算法的聚类精确度和运行效率都得到了提高,预测准确率为98%,召回率为98%,F1值98%,kappa系数高达0.97,远远超过其他集成辨识模型,表明本文模型在汽车用户行为的分辨准确率上得到有效提高。

The driving behavior and operating habits of car users determine whether there is risk in driving,which is of great significance to road traffic safety.In view of this situation,a risk level prediction model based on car users′driving behavior is proposed.The strategy of forgetting the second place is adopted in the SOM neural network algorithm to improve the classification efficiency of the model.The SOM neural network is used to improve K-means clustering analysis method and then the classification of risk behavior of vehicle drivers is realized.After the risk label is obtained by cluster analysis,the XGBoost algorithm is used to identify the user′s risk behavior.The experimental results show that the clustering accuracy and operating efficiency of the improved algorithm is improved;the prediction accuracy rate is 98%,the recall rate is 98%,the F1 value is 98%,and the kappa coefficient is as high as 0.97,far exceeding other integrated identification models.The result shows that the proposed model in this paper is effectively improved in the identification of car user behavior.