论文部分内容阅读
目的分析传染病发病数序列的时空特征,预测山东省传染病乙类、丙类发病情况,提高传染病监测预警能力。方法收集传染病报告信息管理系统山东省法定传染病监测数据资料,应用EViews6.0软件对山东省2010年1月~2014年12月传染病逐月报告发病数构建ARIMA季节乘积模型,并预测2015年1~6月的发病数。结果乙、丙类传染病均呈现一定的周期波动,其中每年的夏季为传染病的高发阶段,秋冬交替时段会有小幅回升;且在山东省17地市有一定差异;ARIMA模型预测与实际的变化趋势基本吻合。结论应用ARIMA对山东省乙类传染病的预测效果优于丙类传染病预测效果,最小的相对误差仅为1.15%,提示该模型可以为传染病的预警提供支持。
Objective To analyze the spatiotemporal features of the number of infectious diseases and predict the incidence of infectious diseases in categories B and C in Shandong Province so as to improve the ability of monitoring and early warning of infectious diseases. Methods Collecting monitoring data of legal infectious diseases in infectious disease reporting information management system in Shandong Province and using EViews6.0 software to build ARIMA seasonal product model of monthly incidence of infectious diseases reported in Shandong Province from January 2010 to December 2014 and forecast 2015 The incidence of 1 to 6 months. Results The epidemics of Category B and Category C showed some periodic fluctuations, of which annual summer was the high-incidence stage of infectious diseases and there was a slight rebound in autumn-winter alternations; there was a certain difference in 17 prefectures and cities in Shandong Province; ARIMA model prediction and actual The trend is basically the same. Conclusions The prediction effect of ARIMA on B infectious diseases in Shandong Province is better than that of C, the minimum relative error is only 1.15%, suggesting that this model can provide support for early warning of infectious diseases.