论文部分内容阅读
研究目的:采用广义估计方程模型对存在时间相关性的事故频次数据进行建模,并与传统广义线性模型的估计效果进行对比。创新要点:通过广义估计方程来考虑事故频次建模中数据的时间相关性,从而提高参数估计准确度以及模型预测精度。研究方法:基于4年高速公路交通事故频次数据,建立考虑时间相关性的广义估计方程以及传统的广义线性模型,并采用统计指标对模型效果进行对比。重要结论:1.事故频次数据样本量对预测精度影响很大;2.广义估计方程能够有效考虑事故频次数据中存在的时间相关性;3.广义估计方程的参数估计比传统广义线性模型更准确,且精度更高。
Research purposes: Generalized estimation equation model was used to model the incident frequency data with time correlation, and compared with the traditional generalized linear model. Innovative Points: Considering the time correlation of data in incident frequency modeling through generalized estimation equations, the accuracy of parameter estimation and the accuracy of model prediction are improved. Research methods: Based on the frequency data of 4-year expressway traffic accidents, a generalized estimation equation considering the time correlation and a generalized generalized linear model are established, and the statistical indexes are used to compare the model results. The main conclusions are as follows: 1. The sample size of accident frequency has a great influence on the prediction accuracy.2. The generalized estimation equation can effectively consider the time correlation in the accident frequency data.3. The parameter estimation of the generalized estimation equation is more accurate than the traditional generalized linear model , And the precision is higher.