We introduce a machine learning framework that uses the differential evolution algorithm in combination with Adam–Bashforth–Moulton method to learn the parameters in a system of variable order fractional differential equations. In this work, we present out developments with regards to taking care of a class of problem: data-driven discovery of system of variable order fractional differential equations. The main advantage of the proposed framework is that it works even if data corresponding to only one of the variables in the system of equations is given. We illustrate the working of our framework on several Examples including modeling the 2014–15 Ebola outbreak in Africa via fractional SEIR (susceptible, exposed, infected, removed) model.

中文翻译：

---

变阶分数阶微分方程组的学习参数

我们介绍了一种机器学习框架，该框架使用差分进化算法结合 Adam–Bashforth–Moulton 方法来学习可变阶分数阶微分方程组中的参数。在这项工作中，我们展示了关于处理一类问题的进展：可变阶分数阶微分方程系统的数据驱动发现。所提出的框架的主要优点是，即使给出的数据仅对应于方程组中的一个变量，它也能起作用。我们在几个示例中说明了我们的框架的工作原理，包括通过部分 SEIR（易感、暴露、感染、移除）模型对 2014-15 年非洲埃博拉疫情进行建模。