General group-based epidemic model for spreading processes on networks: GgroupEM (datasets)

We develop a general group-based continuous-time Markov epidemic model (GgroupEM) framework for any compartmental epidemic model (e.g., susceptible-infected-susceptible, susceptible-infected-recovered, susceptible-exposed-infected-recovered). Here, a group consists of a collection of individual nodes of a network. This model can be used to understand the critical dynamic characteristics of a stochastic epidemic spreading over large complex networks while being informative about the state of groups. Aggregating nodes by groups, the state-space becomes smaller than the one of individual-based approach at the cost of an aggregation error, which is bounded by the well-known isoperimetric inequality. We also develop a mean-field approximation of this framework to reduce the state-space size further. Finally, we extend the GgroupEM to multilayer networks. Individual-based frameworks are in general not computationally efficient. However, the individual-based approach is essential when the objective is to study the local dynamics at the individual level. Therefore, we propose a group-based framework to reduce the computational time of the Individual-based generalized epidemic model framework (GEMF) but retain its advantages.

本研究团队构建了一个通用的基于群体连续时间的马尔可夫病患流行模型（GgroupEM）框架，适用于任何类型的空间流行模型（例如，易感者-感染者-易感者，易感者-感染者-康复者，易感者-暴露者-感染者-康复者）。在此模型中，群体由网络中一系列个体节点集合构成。该模型有助于理解随机流行病在大规模复杂网络中的关键动态特性，同时提供关于群体状态的详细信息。通过将节点按群体进行聚合，状态空间相较于基于个体的方法更小，但需以聚合误差为代价，该误差受已知的等周不等式所限制。此外，我们还开发了该框架的均值场近似，以进一步减小状态空间的大小。最终，我们将GgroupEM扩展至多层网络。基于个体的框架通常在计算效率上不足。然而，当研究目标是探究个体层面的局部动态时，基于个体的方法至关重要。因此，我们提出了一种基于群体的框架，旨在减少基于个体推广的流行模型框架（GEMF）的计算时间，同时保留其优势。