节点文献

道路交通伤害预测方法研究进展

Research Progresses of Predicting Methods of Road Traffic Injuries

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 庞媛媛张徐军崔梦晶顾月陈宇

【Author】 PANG Yuan-yuan;ZHANG Xu-jun;CUI Meng-jing;GU Yue;CHEN Yu;School of Public Health,Southeast University/Injury Prevention Research Institute;

【机构】 东南大学公共卫生学院/伤害预防研究所

【摘要】 道路交通伤害发生率高,造成伤害严重,已成为全球性公共卫生问题。道路交通伤害预测对掌握其未来的发生状况,及时采取相应的措施,具有重要作用。目前常用的预测方法主要有回归分析法、时间序列法、灰色模型法和BP(Back Propagation,BP)神经网络法等。回归分析法主要适用于样本量大、数据波动小、规律性较强的预测;时间序列分析法中应用较多的是自回归滑动平均混合模型(autoregressive integrated moving average model,ARIMA),其对数据的要求不高,具有较好的拟合效果,适用于近期或短期预测;灰色模型法所需样本数据少、原理简单、运算方便,特别适合对具有复杂性、随机性和灰色性特点的道路交通伤害的预测;BP神经网络法则需要较为全面的数据,短期预测精度高,可用于宏观问题的预测。本文就道路交通伤害预测方法进行综述,总结各方法的适用条件及优缺点,为减少和预防道路交通伤害提供依据。

【Abstract】 The road traffic injuries occur at high incidences and cause serious consequences thus have become a serious public health problem worldwide. The existing commonly used predicting methods include regression analysis, time series method, grey model method, and BP(Back Propagation,BP) neural network. Regression analysis is suitable for the prediction of large samples with small fluctuations and a profound regularity. ARIMA(autoregressive integrated moving average model,ARIMA) model is widely applied in time series analysis with acceptable fitting effects, and therefore suitable for the recent or short-term prediction. The Grey model with a simple principle needs smaller samples and is suitable for complex, random and grey data. BP neural network needs more complete data, and the accuracy of its short-term prediction is high. It can be used for macroscopic prediction. This paper reviewed the above mentioned predicting methods of road traffic injuries by describing the applicable conditions, advantages and disadvantages of each method.Comprehensive understanding and careful application of these methods may reduce the incidence of road traffic injuries.

  • 【文献出处】 伤害医学(电子版) ,Injury Medicine(Electronic Edition) , 编辑部邮箱 ,2013年02期
  • 【分类号】U492.8
  • 【下载频次】67
节点文献中: 

本文链接的文献网络图示:

本文的引文网络