数据驱动的具有周期Holling功能反应的云杉蚜虫模型的学习

Data-Driven Learning of a Spruce Budworm Model with Periodic Holling Functional Response

从数据中推断微分方程模型是具有挑战性又重要的课题.本文对云杉蚜虫的实际数据进行分析,通过稀疏回归算法学习构建具有周期Holling功能反应的微分方程模型,模型能很好的与实际数据吻合并能预测云杉蚜虫的周期爆发行为.本工作对探索云杉蚜虫的动力学行为具有指导意义.

Inferring differential equation models from data is a challenging and important subject.In this paper,we analyze the actual data of spruce budworm and learn to construct differential equations with periodic Holling functional response by sparse regression algorithm,which can fit well with the actual data and predict the periodic outbreak behavior of spruce budworm.This work is a guide to explore the kinetic behavior of spruce budworm.