GM(1,1)模型在中国痢疾发病率预测中的应用

Application of GM(1,1)model in forecasting the incidence of dysentery in China

目的分析并预测中国痢疾发病率趋势,为国家公共卫生政策的制定提供参考依据。方法采用灰色GM(1,1)模型对2010—2019年中国痢疾发病率进行拟合,并对2020—2023年中国痢疾发病率进行预测。结果2010—2019年中国痢疾发病率逐年下降,建立的GM(1,1)模型绝对相对误差MAE=4.07%,评价相关误差MAPE=-0.57%,均方误差MSE=0.41,C=0.0907,P=1.0000,模型精度为优秀。模型预测结果显示,2020—2023年中国痢疾各年发病率将会是4.8900/10万、4.2343/10万、3.6666/10万和3.1749/10万。结论构建的预测模型能够很好地预测中国痢疾的流行情况,预测结果显示2020—2023年中国痢疾发病率呈逐年下降趋势,但由于痢疾影响因素存在地区、环境等差异,需要卫生部门因地制宜采取控制策略。

Objective To analyze and predict the incidence trend of dysentery in China,and to provide reference for the formulation of national public health policy.Methods Gray GM(1,1)model was applied to fit the incidence of dysentery in China from 2010 to 2019,and the incidence of dysentery in China from 2020 to 2023 was predicted.Results The incidence of dysentery in China decreased year by year from 2010 to 2019.The GM(1,1)model set up showed relative mean absolute error(MAE)of GM(1,1)model was 4.07%,mean absolute percentage error(MAPE)was evaluated to be-0.57%,mean square error(MSE)satisfied 0.41,C=0.0907,P=1.0000,and the model accuracy was considered to be excellent.The prediction results showed that the incidence of dysentery in China from 2020 to 2023 will be 4.8900/100000,4.2343/100000,3.6666/100000,and 3.1749/100000.Conclusion The established prediction model is eligible to appropriately predict the epidemic situation of dysentery in China,and the prediction results show that the incidence of dysentery in China is decreasing year by year from 2020 to 2023.However,it is necessary for health departments to adopt control strategies according to local conditions due to regional and environmental differences in factors of dysentery.