中南大学学报(医学版)
中南大學學報(醫學版)
중남대학학보(의학판)
JOURNAL OF CENTRAL SOUTH UNIVERSITY (MEDICAL SCIENCES)
2014年
11期
1170-1176
,共7页
手足口病%时间序列%乘积季节自回归积分滑动平均模型
手足口病%時間序列%乘積季節自迴歸積分滑動平均模型
수족구병%시간서렬%승적계절자회귀적분활동평균모형
hand-foot-mouth disease%time series%multiple seasonal autoregressive integrated moving average model
目的:建立长沙市手足口病发病率的乘积季节自回归积分滑动平均模型(autoregressive integrated moving average model,ARIMA),探讨乘积季节ARIMA模型在手足口病疫情预测的可行性。方法:运用EVIEWS 6.0软件对长沙市2008年5月至2013年8月的手足口病发病率资料建立乘积季节ARIMA模型,以2013年9月至2014年2月的发病资料作为模型预测效果的检验样本,最后再用所得到的模型对2014年3月至2014年8月的月发病率进行预测。结果:经过序列平稳化、模型识别以及模型诊断后,建立乘积季节ARIMA模型(1,0,1)×(0,1,1)12,模型拟合度R2=0.81,预测均方根误差为8.29,平均绝对误差为5.83。结论:乘积季节ARIMA模型是一种较好的预测模型,所建模型拟合度较好,能为手足口病的防治工作提供参考。
目的:建立長沙市手足口病髮病率的乘積季節自迴歸積分滑動平均模型(autoregressive integrated moving average model,ARIMA),探討乘積季節ARIMA模型在手足口病疫情預測的可行性。方法:運用EVIEWS 6.0軟件對長沙市2008年5月至2013年8月的手足口病髮病率資料建立乘積季節ARIMA模型,以2013年9月至2014年2月的髮病資料作為模型預測效果的檢驗樣本,最後再用所得到的模型對2014年3月至2014年8月的月髮病率進行預測。結果:經過序列平穩化、模型識彆以及模型診斷後,建立乘積季節ARIMA模型(1,0,1)×(0,1,1)12,模型擬閤度R2=0.81,預測均方根誤差為8.29,平均絕對誤差為5.83。結論:乘積季節ARIMA模型是一種較好的預測模型,所建模型擬閤度較好,能為手足口病的防治工作提供參攷。
목적:건립장사시수족구병발병솔적승적계절자회귀적분활동평균모형(autoregressive integrated moving average model,ARIMA),탐토승적계절ARIMA모형재수족구병역정예측적가행성。방법:운용EVIEWS 6.0연건대장사시2008년5월지2013년8월적수족구병발병솔자료건립승적계절ARIMA모형,이2013년9월지2014년2월적발병자료작위모형예측효과적검험양본,최후재용소득도적모형대2014년3월지2014년8월적월발병솔진행예측。결과:경과서렬평은화、모형식별이급모형진단후,건립승적계절ARIMA모형(1,0,1)×(0,1,1)12,모형의합도R2=0.81,예측균방근오차위8.29,평균절대오차위5.83。결론:승적계절ARIMA모형시일충교호적예측모형,소건모형의합도교호,능위수족구병적방치공작제공삼고。
Objective: To establish multiple seasonal autoregressive integrated moving average model (ARIMA) according to the hand-foot-mouth disease incidence in Changsha, and to explore the feasibility of the multiple seasonal ARIMA in predicting the hand-foot-mouth disease incidence. Methods: EVIEWS 6.0 was used to establish multiple seasonal ARIMA according to the hand-foot-mouth disease incidence from May 2008 to August 2013 in Changsha, and the data of the hand-foot-mouth disease incidence from September 2013 to February 2014 were served as the examinedsamples of the multiple seasonal ARIMA, then the errors were compared between the forecasted incidence and the real value. Finally, the incidence of hand-foot-mouth disease from March 2014 to August 2014 was predicted by the model. Results: Atfer the data sequence was handled by smooth sequence, model identiifcation and model diagnosis, the multiple seasonal ARIMA (1, 0, 1)×(0, 1, 1)12 was established. The R2 value of the model iftting degree was 0.81, the root mean square prediction error was 8.29 and the mean absolute error was 5.83. Conclusion: hTe multiple seasonal ARIMA is a good prediction model, and the iftting degree is good. It can provide reference for the prevention and control work in hand-foot-mouth disease.