SARIMA模型在泰安市流行性腮腺炎发病预测中的应用  

Application of SARIMA model in predicting the incidence of mumps in Tai'an city

作　　者：李楠 国青 徐丽莎[1] 明明[1] 石艳艳[1] LI Nan;GUO Qing;XU Li-sha;MING Ming;SHI Yan-yan(Immunization Planning Department,Taian City Center for Disease Control and Prevention,Shandong 271000,China)

机构地区：[1]泰安市疾病预防控制中心免疫规划科,山东271000

出　　处：《预防医学论坛》2024年第8期593-596,共4页Preventive Medicine Tribune

摘　　要：目的 构建流行性腮腺炎(流腮)发病数的季节性差分自回归移动平均(SARIMA)模型,并对泰安市流腮发病情况进行预测,为流腮防控依据。方法 流腮发病数据资料来源于中国疾病预防控制信息系统中监测报告管理系统,按照现住地址、发病日期收集2010年1月-2021年6月泰安市流腮发病数据资料。采用SPSS 21.0根据2010-2020年的数据建立最优SARIMA模型,利用该模型对2021年1-6月的流腮月发病数进行预测,评价模型的拟合效果,并对2021年7-12月流腮发病情况进行预测。结果 泰安市2010-2020年流腮月发病趋势具有一定的波动性,存在季节周期性特点。SARIMA(0,1,2)(1,1,0)_(12)模型为最优预测模型,该模型拟合度R~2=0.773,贝叶斯信息准则(BIC)值为7.162。Q=9.877,P>0.05,该模型残差为白噪声序列,所构建的模型较为合理。结论 SARIMA(0,1,2)(1,1,0)_(12)模型拟合效果较好,可用于泰安市流腮月发病趋势的短期预测。Objective To establish a Seasonal Autoregressive Integrated Moving Average(SARIMA)model for the incidence of mumps and forecast the mumps incidence in Tai'an city,so as to provide a basis for mumps prevention and control.Methods The mumps incidence data were sourced from the Monitoring and Reporting Management System of China's Disease Prevention and Control Information System.Data on mumps incidence in Tai'an city from January 2010 to June 2021 were collected based on current residential address and date of onset.Using SPSS 21.0,the optimal SARIMA model was established based on data from 2010 to 2020.This model was then used to predict the monthly incidence of mumps for January to June 2021,evaluate the model's fitting effect,and forecast the mumps incidence for July to December 2021.Results The monthly incidence trend of mumps in Tai'an city from 2010 to 2020 exhibited certain fluctuations and seasonal periodicity.The SARIMA(0,1,2)(1,1,0)_(12)model was identified as the optimal prediction model,with a goodness-of-fit R~2=0.773 and a Bayesian Information Criterion(BIC)value of 7.162.The Q-statistic was 9.877,and P>0.05,indicating that the model residuals were white noise sequences,suggesting that the constructed model was reasonable.Conclusion The SARIMA(0,1,2)(1,1,0)12 model exhibits good fitting performance and can be used for short-term forecasting of the monthly incidence trend of mumps in Tai'an city.

关 键 词：流行性腮腺炎 时间序列分析 SARIMA模型 预测 

分 类 号：R512.1[医药卫生—内科学]