ARIMA模型预测深圳宝安区流感样病例发病率的可行性

Feasibility of Incidence of influenza-like illness (ILI) in Bao'an district predicted by ARIMA model

目的建立预测宝安区流感样病例发病率的ARIMA模型,并验证其可行性。方法应用SPSS 19.0软件对2010年1月~2015年6月宝安区流感样病例每月发病率进行模型拟合,以2015年7~9月每月流感样病例发病率作为考核样本评价模型预测效果。结果 ARIMA(2,1,0)(0,1,1)12拟合2010年1月~2015年6月宝安区每月新增感染率的变动趋势较为理想,实际值均在预测值的95%可信区间内,预测结果与实际值相对误差最大值为20.94%,最小为3.05%。结论 ARIMA(2,1,0)(0,1,1)12模型,该模型在宝安流感样病例发病率短期趋势的预测与实际发病率吻合,可考虑纳入周期性自变量或者选用其他模型加强预测精度。

Objective To verify the feasibility of the incidences of influenza-like illness （ILI） in Bao＇an district predicted by ARIMA model. Methods Using SPSS 19.0 to build the ARIMA model and to fit the incidences of influenza-like cases from 2010.1 to 2015.6 in Bao＇an district ,to predict the incidences form 2015.6 to 2015.9,and to evaluate the prediction effects. ResuRs The best fitting model is ARIMA （2,1,0） （0,1,1）12model, Actual values are within the 95% confidence interval of predictive value. Relative errors between prediction results and actual value, the maximum is 20.94%, the minimum is 3.05%. Conclusions The predictions of ARIMA （2,1,0） （0,1,1h2 model agree well with the actual incidence of influenza-like illness in Bao＇an district in a short time. Including periodic independent variables or trying other models can strengthen the accuracy.