ARIMA乘积季节模型在流行性腮腺炎发病率预测中的应用

目的构建时间序列分析ARIMA乘积季节性模型，预测武汉市2013年流行性腮腺炎发病趋势，探讨该模型在流行性腮腺炎发病率预测中的应用。方法利用中国疾病监测信息报告系统2005年1月至2012年12月武汉市流行性腮腺炎月发病资料，运用PASWStatistics18．0专家建模器，考虑季节性因素建立ARIMA预测模型，并将所建模型对2013年流行性腮腺炎月发病率进行预测。结果模型残差序列为白噪声，ARIMA（0，1，0）（0，1，1）。：模型构建理想，能较好地拟合时间序列，2013年流行性腮腺炎月发病高峰在5～7月，至11、12月出现次要高峰。结论专家建模器构建乘积季节性模型拟合效果较好，但在实际工作中要充分考虑流行性腮腺炎各项防控措施推行等影响因素，综合分析，为防控工作提出理论与实证建议。

The mumps incidence prediction by multiple seasonal ARIMA model

Objective To establish a multiple seasonal AILIMA model, a time series analysis, for predicting the mumps incidence trend in Wuhan in 2013, and explore the application of ARIMA model for mumps incidence prediction. Methods Reported mumps monthly incidence in Wuhan from January 2005 to December 2012 were collected from China Information System for Disease Control and Prevention. A multiple seasonal ARIMA model was established using the ＂Expert Modeler＂ in PASW Statistics 18.0 to predict mumps monthly incidence in 2013 in Wuhan. Results The residual series of the ARIMA （0,1,0） （ 0,1,1 ） t2 model established were white noise, indicating that the ARIMA model was credible and could fit time series well. According to the ARIMA model,the peak of the mumps monthly incidence in 2013 would appear in May, June and July, and the secondary peak would appear in November and December. Conclusion The multiple seasonal ARIMA model was successfully established using the Expert Modeler and could provide a good fit. However,the factors related to the measures of prevention and control of mumps should be fully considered and comprehensively analyzed in the practical work before putting forward theoretical and empirical suggestions for prevention and control work.