论文部分内容阅读
道路交通事故鉴定对于交通事故责任认定和法庭举证具有重要作用.采集某市近三年125起交通事故案例数据,建立机动车、自行车、人体总计73个调查变量,运用SPSS软件开展骑推行事故鉴定研究.通过相关性分析发现车座、机动车类型等6个变量与鉴定结论相关.开展线性判别分析研究,结果表明直接对相关变量进行判别分析可以快速获得最佳判别效果.当函数选入的自变量为车座时,交叉验证准确率最高可达72.8%,说明采用数据挖掘的方法来鉴别交通事故中的行为方式具有一定可行性.
Identification of road traffic accidents plays an important role in the identification of liability and forensic evidence.According to the data of 125 cases of traffic accidents collected in a certain city in recent three years, 73 survey variables of motor vehicles, bicycles and human bodies are set up, and SPSS software is used to carry out the identification of riding and driving accidents The correlation analysis shows that six variables such as vehicle seat and vehicle type are related to the conclusion of the appraisal.The linear discriminant analysis is carried out and the results show that the discriminant analysis of the related variables can quickly obtain the best discriminant effect.When the function is selected When the independent variable is a car seat, the accuracy of cross-validation can reach as high as 72.8%, which shows that it is feasible to use data mining method to identify the behavior in traffic accidents.