细菌性痢疾自回归滑动平均和非线性自回归组合模型预测研究

发布时间：2018-08-18 09:40

【摘要】：目的探讨单纯自回归滑动平均(autoregressive integrated moving average,ARIMA)模型与ARIMA和非线性自回归(nonlinear autoregressive,NAR)组合模型在细菌性痢疾预测中的应用。方法利用江苏省2004年1月至2015年2月的细菌性痢疾数据作为拟合样本,以2015年3月至2016年5月的数据作为预测样本;建立的模型分别为单纯ARIMA模型和ARIMA-NAR组合模型,然后根据2个模型的平均绝对误差(mean absolute error,MAE)、均方误差(mean square error,MSE)和平均绝对百分比误差(mean absolute percentage error,MAPE)比较模型的效果,其值越小模型效果越好。结果在模型的拟合阶段,单纯ARIMA模型的MAE、MSE和MAPE分别为0.177 5、0.081 4和0.184 7,ARIMA-NAR组合模型分别为0.094 1、0.029 5和0.104 6。在模型的预测阶段,单纯ARIMA模型的MAE、MSE和MAPE也分别大于ARIMA-NAR组合模型。结论 ARIMA-NAR组合模型对于江苏省细菌性痢疾发病率时间序列的预测效果优于单纯ARIMA模型。建议尝试使用ARIMA-NAR组合模型预测细菌性痢疾的发病率。
[Abstract]:Objective to investigate the application of simple autoregressive moving average (autoregressive integrated moving averagea (ARIMA) model, ARIMA and nonlinear autoregressive model in prediction of bacillary dysentery. Methods the data of bacillary dysentery from January 2004 to February 2015 in Jiangsu Province were used as fitting samples and the data from March 2015 to May 2016 were used as prediction samples. The established models were simple ARIMA model and ARIMA-NAR combined model, respectively. Then, according to the mean absolute error (mean absolute), mean square error (mean square) and mean absolute error (mean absolute percentage) of the two models, the model is compared. The smaller the model value is, the better the model effect is. Results in the fitting stage of the model, the mae MSE and MAPE of the simple ARIMA model were 0.177 5 / 0.081 4 and 0.184 7 / 7 ARIMA-NAR respectively, which were 0.094 1 / 0.029 5 and 0.104 6 respectively. In the prediction stage of the model, the MAE MSE and MAPE of the simple ARIMA model are also larger than that of the ARIMA-NAR combination model, respectively. Conclusion ARIMA-NAR combined model is superior to ARIMA model in predicting the time series of incidence of bacillary dysentery in Jiangsu Province. It is suggested that ARIMA-NAR model be used to predict the incidence of bacillary dysentery.
【作者单位】：江南大学无锡医学院流行病学与卫生统计学教研室;
【基金】：江南大学自主科研青年基金(JUSRP11569) 江南大学公共卫生研究中心项目(JUPH201508)~~
【分类号】：R516.4

【相似文献】