Objective: To predict the incidence of local malaria of Hubei Province applying the Autoregressive Integrated Moving Average model (ARIMA).
Methods: SPSS 13.0 software was applied to construct the ARIMA model based on the monthly local malaria incidence in Hubei Province from 2004 to 2009. The local malaria incidence data of 2010 were used for model validation and evaluation.
Results: The model of ARIMA (1, 1, 1) (1, 1, 0)12 was tested as relatively the best optimal with the AIC of 76.085 and SBC of 84.395. All the actual incidence data were in the range of 95% CI of predicted value of the model. The prediction effect of the model was acceptable.
Conclusions: The ARIMA model could effectively fit and predict the incidence of local malaria of Hubei Province.
［摘要］ 目的 应用自回归求和移动平均模型 (Autoregressive Integrated Moving Average Model, ARIMA) 进行湖北省本地 疟疾发病率预测。方法 应用SPSS 13.0软件对2004-2009年湖北省本地疟疾发病率构建ARIMA模型, 并以2010年发 病率数据检验模型, 评价模型拟合及预测效果。结果 经检验确认ARIMA (1, 1, 1) (1, 1, 0) 12模型拟合效果相对最 优, AIC=76.085, SBC=84.395, 发病率实际值均在预测值的95%可信区间内, 表明模型预测效果较好。 结论ARIMA模 型可对湖北省本地疟疾发病率进行较好的拟合和预测。.
Keywords: ARIMA model; Incidence; Malaria; Prediction; Time series.