基于自回归滑动平均混合模型的安丘市肾综合征出血热发病率预测研究  

Study on incidence prediction of hemorrhagic fever of renal syndrome (HFRS) by autoregressive integrated moving average (ARIMA) model,Anqiu city

作　　者：李方友 孟凡强 禹长兰 于世成 LI Fang-you;MENG Fan-qiang;YU Chang-lan;YU Shi-cheng(Anqiu City Center for Disease Control and Prevention,Anqiu,Shandong,262100,China)

机构地区：[1]安丘市疾病预防控制中心,山东安丘262100 [2]安丘市医养健康产业发展中心,山东安丘262100

出　　处：《预防医学论坛》2020年第2期109-112,共4页Preventive Medicine Tribune

摘　　要：目的基于自回归滑动平均混合(ARIMA)模型对安丘市肾综合征出血热(HFRS)的发病率进行时间序列分析和预测,为制定HFRS防治策略提供科学依据。方法收集安丘市1977年1月1日至2019年8月31日安丘市HFRS发病资料,按照年、月进行分类汇总,通过SPSS 19.0软件拟合ARIMA模型,分别预测2019~2020年和2019年1~8月的发病率。结果年发病率预测模型最终拟合为ARIMA(0,1,0)模型,预测2019、2020年的年发病率分别为2.52/10万和1.56/10万。月发病率预测模型最终拟合为ARIMA(0,1,2)(0,1,1)12。结论 ARIMA模型能较好模拟安丘市HFRS的发病趋势。Objective To conduct time series analysis and prediction of hemorrhagic fever of renal syndrome(HFRS) incidence by ARIMA model in Anqiu city,so as to provide scientific evidence for developing the prevention and control strategy of HFRS. Methods The surveillance data of HFRS from January 1977 to August 2019 in Anqiu city were collected and sorted by year and month for the establishment of ARIMA model by SPSS 19.0,ARIMA model was fitted under time series to predict the HFRS annual incidence of 2019-2020 and monthly incidence from January to August in 2019. Results The annual incidence prediction model was finally fitted as ARIMA(0,1,0) model,and the predictive annual incidence in 2019 and 2020 was 2.52/10~5 and 1.56/10~5 respectively.The ARIMA(0,1,2)(0,1,1)12 model had the best fitting effect in the prediction of monthly incidence. Conclusion ARIMA model could be used to successfully predict the incidence of HFRS in Anqiu city.

关 键 词：肾综合征出血热 时间序列 自回归滑动平均混合模型 预测 

分 类 号：R512.8[医药卫生—内科学]