指数平滑法与ARIMA模型在湖北省丙型病毒性肝炎发病预测中的应用

发布时间：2018-02-11 17:48

本文关键词： 丙型病毒性肝炎指数平滑法 ARIMA模型发病率预测　出处：《中国疫苗和免疫》2017年03期 　论文类型：期刊论文

【摘要】：目的分析指数平滑法和ARIMA模型对湖北省丙型病毒性肝炎(丙肝)发病率的预测效能。方法选取湖北省2004年1月-2014年12月丙肝发病数据作为指数平滑法和ARIMA模型的建模部分,2015年1-12月丙肝发病数据作为模型验证部分,比较两种模型的拟合和预测效果优劣。结果 2004-2015年湖北省丙肝报告发病率呈现明显上升趋势,ARIMA(0,1,1)×(0,1,1)_(12)模型拟合的均方根误差(RMSE)为0.076,2015年1-12月预测值年平均相对误差为1.67%;运用指数平滑法构建的模型是Winters加法模型,其拟合结果 RMSE为0.086,预测2015年1-12月的年平均相对误差为2.12%。结论 ARIMA模型拟合效果较好,预测精度更高,可应用于丙肝疫情的预测和预警。
[Abstract]:Objective to analyze the predictive efficacy of exponential smoothing method and ARIMA model for the incidence of hepatitis C in Hubei Province. Methods the data of hepatitis C from January 2004 to December 2014 in Hubei Province were selected as exponential smoothing method and ARIMA model. In the modeling part, from January to December of 2015, the incidence data of hepatitis C were used as the verification part of the model. Results the reported incidence rate of hepatitis C in Hubei Province from 2004 to 2015 showed a significant upward trend. The RMSE of the model fitting was 0.076, and the average annual relative error of the predicted value in January-December of 2015 was 1.6775, and the mean square error (RMSE) of the two models was 0.076, and the mean relative error of the predicted value in 2015 was 1.677.Results the incidence rate of hepatitis C in Hubei Province showed an increasing trend from 2004 to 2015. The mean square root error (RMSE) of the model fitted was 0.076. The model constructed by exponential smoothing method is Winters addition model. The fitting result RMSE is 0.086 and the average annual relative error between 2015 and December is 2.12.Conclusion the fitting effect of ARIMA model is better and the prediction accuracy is higher. It can be applied to the prediction and early warning of hepatitis C epidemic situation.
【作者单位】：江汉大学医学院;湖北省疾病预防控制中心;
【分类号】：R181.3;R512.63

【相似文献】