摘要
目的应用时间序列模型中较有代表性的3种方法对江苏省伤寒副伤寒疫情进行分析预测,并探讨3种方法的优缺点及适用性,为传染病疫情预测分析及防治提供理论依据。方法以1980年至2008年江苏省伤寒副伤寒发病资料分别建立合适的乘积季节模型、指数平滑模型和状态空间模型,以2009年的发病资料作为模型预测效果的考核样本,对预测结果进行深入分析和评价,探讨对疫情进行预测的方法、思路和应该注意的问题。结果江苏省伤寒副伤寒的发病趋势将在以后一段时间内呈平稳下降的趋势,会存在少许波动。乘积季节模型的预测结果与实际情况最相符,指数平滑法的预测效果较差,状态空间模型也有一定的实用性。结论用时间序列模型对伤寒副伤寒发病情况的拟合结果较好,尤其是乘积季节模型,预测效果良好,可为伤寒副伤寒的防治提供新的信息和思路。
Objective To research the application of three representative methods of time series model in forecasting infectious disease, and explore the relative merits and applicability of the three models. Methods Models were fitted by the historical data of typhoid and paratyphoid of Jiangsu Province. Using the established models to predict the morbidity and then to analyze the forecasting results, discuss the problems we need to pay attention and the way to improve the veracity and practicability of forecasting. Results The trends of incidence of typhoid and paratyphoid would continue to decline steadily in the near future, and at the same time, the product season model was thought to be the most appropriate of the three, the exponential smoothing model was not quite suitable, the state space method had its certain applicability. Conclusions The time series model is practical for predicting typhoid and paratyphoid in Jiangsu Province, especially the product season model, for it can provide exceptive information and ideas to prevent and cure the infectious disease like typhoid and paratyphoid.
出处
《河南预防医学杂志》
2014年第4期244-248,共5页
Henan Journal of Preventive Medicine
关键词
乘积季节模型
指数平滑法
状态空间模型
伤寒副伤寒
预测
product season model
exponential smoothing model
state space method
typhoid and paratyphoid
forecasting
