RAPID: COVID-19 Response Support: Building Synthetic Multi-scale Networks

新奇的2019年冠状病毒(新冠肺炎)已经在全球造成了前所未有的社会、经济和健康影响。该项目将开发综合的全球多尺度社会联系网络。合成的但现实的社交联系网络可以捕捉到个人或社区层面的人类互动。这些网络可以与基于代理的模型结合使用，以模拟正在进行的新冠肺炎大流行。这些模拟反过来可以用来设计和评估各种干预措施，以平衡健康益处与社会和经济成本。将向科学界提供数据。PIS还将与其他研究小组合作，并继续与其他联邦和州机构建立伙伴关系，以支持他们的应对努力。开发合成的社交网络是一个在统计学和算法上都具有挑战性的问题。该项目将使用机器学习和数据驱动建模技术的组合来综合两类合成社会联系网络--基于补丁的元群体网络和个性化的合成社会联系群体和网络。在可用数据稀疏且有噪声的情况下，对这种数据驱动的机械建模方法的需求已变得非常明显。该项目将对算法及其产生的合成网络进行详细的统计分析。这包括进行全球敏感性分析的方法，以及根据网络结构对结果的不确定性进行量化的方法。这种资源的众多用途之一是支持基于个体和基于元种群的流行病传播模拟模型，特别是新冠肺炎。除了支持持续的新冠肺炎爆发，这些人工合成的社交联系网络将有助于应对其他流行病。PIS计划将这些数据提供给全球研究界，以便世界各地的研究人员可以立即使用它来评估大流行和各自地区的应对工作。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

The Novel 2019 Coronavirus (COVID-19) has already caused unprecedented global social, economic, and health impact. This project will develop synthetic global multi-scale social contact networks. The synthetic but realistic social contact networks can capture human interactions either at an individual or community level. The networks can be used in conjunction with agent-based models to simulate the ongoing COVID-19 pandemic. The simulations can in-turn be used to design and assess various interventions that balance health benefits with social and economic costs. Data will be made available to the scientific community. The PIs will also work with other research groups and continue their partnership with other federal and state agencies to support their response efforts. Developing synthetic social contact networks is a statistically and algorithmically challenging problem. This project will synthesize ensembles of two classes of synthetic social contact networks -- patch-based meta-population networks and individualized synthetic social contact populations and networks using a combination of machine learning and data driven modeling techniques. The need for such data driven mechanistic modeling methods has become abundantly clear in regimes when the available data is sparse and noisy. The project will undertake a detailed statistical analysis of the algorithms and the synthetic networks they produce. This includes methods to conduct global sensitivity analysis and methods to quantify the uncertainty in the outcomes as a function of the network structure. One of the many uses of this resource, is to support individual-based as well as meta-population-based simulation models for epidemic spread in general, and COVID-19 in particular. Beyond supporting ongoing COVID-19 outbreaks, these synthetic social contact networks will be useful in responding to other epidemics. The PIs plan to make this data available to the global research community so that researchers around the world can immediately use it to assess the pandemic and the response efforts in their respective regions.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.