ARIMA模型在安徽省流行性感冒发病预测中的应用

Pathogenesis prediction of influenza in Anhui Province by the ARIMA Model

目的探讨ARIMA模型在安徽省流感发病预测应用中的可行性并开展预测,为流感的防控提供理论依据。方法收集安徽省2005-2012年流感发病资料建立数据库,运用SPSS 13.0对逐月发病率进行ARIMA建模与拟合,在数据平稳化、定阶和参数估计后,建立ARIMA数学模型,并利用模型对2013年发病情况开展预测。结果建立ARIMA(2,2,1)(2,2,0)12模型,预测结果符合实际发病趋势,该模型具有实用性。结论 ARIMA模型可用于流感发病率在短期内变化趋势的拟合与预测。

Objective To explore the application of time series analysis on influenza and confirm its feasibility and applicability.Methods Auto-regressive integrated moving average（ARIMA） model was fitted with data of monthly influenza incidence from January2005 to June 2012,and the monthly incidence of 2013 was predicted and evaluated.Results The model of ARIMA（2,2,1）（2,2,0）12was established and used to verify its usability due to predictive results being similar to the actual incidence.Conclusion The ARIMA model can be used to fit trends of influenza incidence on a time series and forecast changing trends within a short period.