数据驱动的突发疫情传染扩散参数动态更新策略

在对各类重大突发疫情进行建模拟合时,参数取值一直是困扰众多学者的重要现实难题,现有研究大多参考相关文献或结合医学实验选取某一固定参数.为克服这种固定参数取值的局限性,基于数据驱动的逆向思维,借助欧拉差分变换和线性方程组解的特性,构建一种疫情传染扩散参数动态更新策略,可以帮助决策者结合疫情实时更新的状态数据,反推计算最佳的疫情传播扩散参数.以武汉新冠肺炎疫情相关数据进行算例测试,结果表明,结合所设计的参数动态更新策略,能够有效地提升重大突发疫情演化预测的准确性,这对于政府应急资源的精准配置具有重要的决策支持作用.

A dynamic design strategy of epidemic spreading parameters based on data driven

When modeling and fitting various kinds of epidemic outbreaks, the value of parameters has always been an important practical problem for many scholars. In the existing studies, most of the authors select a fixed parameter by referring to the relevant literature or combined with medical experiments. With the help of Euler difference transformation and the characteristics of the solution of linear equations, we innovatively propose a dynamic update strategy of epidemic diffusion parameters based on data-driven in this study in order to overcome the above limitation. The method can help decision-makers to calculate the optimal parameters of epidemic spread by combining the real-time update data. A case study is conducted with the COVID-19 data of Wuhan. The results show that the dynamic parameter update strategy designed in this paper can effectively improve the accuracy of the evolution prediction of epidemic outbreaks, which provides an important decision support for the accurate allocation of government emergency resources.