多种数据模型在手足口病发病预测的应用探讨

Application of various data models in predicting the incidence of hand,foot and mouth disease

目的应用多种数学模型拟合出手足口病发病最优模型,预测手足口病发病趋势,为疾病防控提供科学依据。方法采用SPSS 18. 0软件建立差分自回归移动平均模型(autoregressive integrated moving average,ARIMA模型)、差分自回归移动平均模型与多层感知神经网络组合模型(autoregressive integrated moving average and multilayer perceptron,ARIMA-MLP模型)和差分自回归移动平均模型与径向基函数神经网络组合模型(autoregressive integrated moving average and radial basis function,ARIMA-RBF模型),分别对手足口病发病情况进行拟合,通过对三种模型比较,找到预测最优模型。结果 ARIMA模型的拟合度R2和平均绝对误差MAE值分别为0. 725、2. 672,ARIMA-MLP模型为0. 724、2. 672,ARIMA-RBF模型为0. 801、2. 206。结论 ARIMA-RBF模型的拟合度最大、平均绝对误差最小模型,拟合效果优于其他两种模型。

Objective To predict the incidence trend of hand,foot and mouth disease（ HFMD） through an optimal prediction model of HFMD incidence fitted by various mathematical models,and to provide a scientific basis for HFMD prevention and control. Methods SPSS18.0 software was employed to establish ARIMA,ARIMA-MLP and ARIMA-RBF models,and then the models were respectively used to fit the incidence of HFMD. The optimal prediction model was found out by comparing the three models. Results The goodness of fit R2 and MAE（ mean absolute error） values of ARIMA,ARIMA-MLP and ARIMA-RBF models were 0.725 and 2.672,0.724 and 2.672,0.801 and 2.206 respectively. Conclusions ARIMA-RBF model has maximum R2 and minimum MAE,and its fitting efficiency is superior to those of two other models.