文献类型：期刊文献

中文题名：定西市安定区手足口病SARIMA预测模型研究

英文题名：Research of SARIMA prediction model for hand-foot-mouth disease in Anding District of Dingxi City

作者：马丽娜[1];刘思强[1];陆小梅[2]

第一作者：马丽娜

机构：[1]甘肃中医药大学定西校区医学教学部检验教研室,甘肃定西743000;[2]定西市安定区疾病控制中心

第一机构：甘肃中医药大学定西校区

年份：2021

卷号：37

期号：1

起止页码：59

中文期刊名：中国公共卫生管理

外文期刊名：Chinese Journal of Public Health Management

收录：国家哲学社会科学学术期刊数据库

基金：甘肃省高等学校创新能力提升项目——甘肃定西地区手足口病预测模型的构建(2019B-230);甘肃中医药大学定西校区项目(2019XJYB03)。

语种：中文

中文关键词：手足口病;SARIMA模型;发病率;预测

外文关键词：hand-foot-mouth disease;SARIMA model;incidence;prediction

摘要：目的构建自回归求和移动平均模型(SARIMA),用于定西市安定区手足口病发病率的预测,分析该模型在手足口病预测中的应用。方法根据2008年1月—2019年6月手足口病月发病率构建SARIMA模型,2019年7—12月手足口病发病率为验证数据,验证模型预测效果。结果SARIMA(1,0,1)×(0,1,1)12为最优模型,其平稳R2为0.754,均方根误差为6.540,正态化的BIC值为3.911。预测结果显示:2019年7-12月手足口病月发病率预测最小相对误差为0.51%,最大相对误差为26.18%,平均相对误差为16.01%,实际值均在预测值的95%CI以内。结论SARIMA(1,0,1)×(0,1,1)12模型能较好地拟合安定区手足口病月发病情况,在实际应用中,SARI?MA模型适用于短期预测,同时本模型预测值的95%CI宽度偏大,所以在后期的工作中,应对模型进一步优化。
Objective A SARIMA model was established to predict the incidence trend of hand-foot-mouth disease in Anding District,Dingxi City.Methods The SARIMA model was established for the monthly incidence of hand-footmouth disease from January,2008 to June,2019,and the incidence of hand-foot-mouth disease from July to December,2019 was a validation data to verify the predictive effect of the model.Results SARIMA(1,0,1)×(0,1,1)12 model was the optimal model,with Smooth R2 of 0.754,root-mean square error of 6.540,and normalized BIC value of 3.911.The pre?diction results showed that the minimum relative error was 0.51%,the maximum relative error was 26.18%and the mean relative error was 16.01%between the predicted and actual monthly incidence of hand-foot-mouth disease from July to December,2019,and the actual value was within the 95%confidence interval of the predicted value.Conclusion The SARIMA model can better predict the incidence trend of hand-foot-mouth disease in Anding District.However,the SARIMA model is suitable for short-term prediction and the predicted value of the 95%confidence interval with the model is too wide.So in the later work,the prediction effect of the model needs to be further optimized.