ARIMA-BPNN组合预测模型在流感发病率预测中的应用

发布时间：2018-02-26 10:02

本文关键词： ARIMA模型 ARIMA-BPNN组合模型预测流感发病率　出处：《中国卫生统计》2014年01期 　论文类型：期刊论文

【摘要】：目的阐述ARIMA-BPNN组合模型预测流感发病率的方法和步骤,探讨其在流感发病率预测中的应用。方法利用河南省2004年1月-2010年12月的流感疫情数据作为训练集,建立ARIMA模型和ARIMA-BPNN组合模型,选取2011年1月-12月的疫情数据作为检验集,评价模型的预测效能。结果 ARIMA(3,0,0)模型和ARIMA-BPNN组合模型预测值的平均绝对误差及平均误差绝对率分别为1.438、27.65%和0.029、0.43%,ARIMA-BPNN组合模型的预测效能优于ARIMA模型。结论 ARIMA-BPNN组合模型能有效模拟、预测流感的发病疫情,具有较好的推广应用价值。
[Abstract]:Objective to describe the methods and steps of ARIMA-BPNN combined model for influenza incidence prediction, and to explore its application in influenza incidence prediction. Methods the influenza epidemic data from January 2004 to December 2010 in Henan Province were used as training set. The ARIMA model and ARIMA-BPNN combination model were established, and the epidemic data from January 2011 to December were selected as the test set. Results the average absolute error and the average absolute rate of the predicted values of Arima model and ARIMA-BPNN combination model were 1.438% 27.65% and 0.029% 0.43% respectively. The prediction efficiency of the combined model was better than that of the ARIMA model. Conclusion the combined ARIMA-BPNN model can be used to simulate effectively. It is of great value to predict the incidence and epidemic situation of influenza.
【作者单位】：郑州大学公共卫生学院流行病与卫生统计学系;郑州大学信息工程学院软件工程系;河南省医学科学院;
【基金】：国家自然科学基金(No:81001293) 河南省教育厅自然科学基金(2010B330004) 郑州大学全国大学生创新创业训练计划项目(2012cxsy130&2011cxsy173)
【分类号】：R184

【参考文献】