GM(1,1)残差修正模型在中国淋病发病率预测中的应用

Application of GM(1,1) residual correction model in predicting the incidence of gonorrhea in China

目的 探讨GM(1,1)残差修正模型在淋病发病率预测中的适用性,并分析我国淋病发病率趋势,提出防控建议。方法 基于2012-2019年国家卫生部门公开数据,采用描述性统计学的方法对发展趋势进行分析,利用Python软件建立我国淋病发病率预测的GM(1,1)模型及其残差修正模型。结果 2012-2019年,我国淋病发病率呈倒“v”型波动变化,发病率顺位移至我国传染病的第4顺位;研究直接构建的GM(1,1)模型精度评价为不合格,经4次修正得到模型精度评价为优秀的预测模型,参数C=0.290 3,P=1.000 0,最终模型预测结果显示,2020-2024年我国淋病发病率预测区间分别为(9.199 3/100 000-9.870 5/100 000)、(9.706 3/100 000-10.414 5/100 000)、(10.288 4/100 000-11.039 1/100 000)、(10.950 1/100 000-11.749 1/100 000)和(11.697 2/100 000-12.550 7/100 000)。结论 GM(1,1)残差修正模型能够很好弥补GM(1,1)模型存在的不足,可用于淋病发病率预测。

Objective To explore the applicability of GM(1,1) residual correction model in predicting the incidence of gonorrhea, and analyze its incidence trend, so as to put forward suggestions for prevention and control of gonorrhea in China.Methods Based on the public data of the National Health Departments from 2012 to 2019,the development trend was analyzed by descriptive statistics.Python software was used to establish the GM(1,1) model and its residual correction model for predicting the incidence of gonorrhea in China.Results The incidence of gonorrhea in China fluctuated in an inverted “v” shape from 2012 to 2019,and the incidence rate shifted to the fourth among the infectious diseases.The accuracy evaluation of the directly constructed GM(1,1) model was unqualified.After four revisions, the model accuracy was evaluated as excellent, with parameters C=0.290 3,P=1.000 0.The prediction results of the excellent model showed that the forecast intervals of gonorrhea incidence in China were(9.199 3/100 000-9.870 5/100 000),(9.706 3/100 000-10.414 5/100 000),(10.288 4/100 000-11.039 1/100 000),(10.950 1/100 000-11.749 1/100 000) and(11.697 2/100 000-12.550 7/100 000) from 2020 to 2024.Conclusion The GM(1,1) residual error correction model can make up for the shortcomings of the GM(1,1) model, and can be used to predict the incidence of gonorrhea.