COVID 19: RAPID: Informed Decision Making for Pandemic Management

该提案的主要目标是开发一套工具，为决策者提供信息和建议，以控制大规模全球流行病。需要解决的具体研究问题包括：(1)对未来几天/几周/几个月的大流行状况作出可靠的预测。(2)估计政府措施对流行病动态的影响。(3)设计有效控制大流行的方案，防止医疗系统超载。有效地管理大规模流行病，这些流行病可能会使全国范围内的卫生保健系统不堪重负，这肯定会导致人命损失的显著增加，包括与引起流行病的特定病原体无关的死亡。该项目将为开发流行病估计和控制的可操作模型的基础研究做出贡献。具体而言，重点是适合于(i)估计模型参数、（ii）估计流行病状况和（iii）计算最佳干预政策的模型；根据流行病演变过程中可用的相对较小的数据集，这三者都具有高度的可信度。这项研究包括数据驱动技术，以确定非药物社会措施对流行病管理的影响。这一领域的一个关键挑战是可靠地推断社会措施及其对流行病动态的影响之间的因果关系。将制定方法，确定干预政策，尽量减少人员生命损失和经济影响，而不会使卫生保健系统不堪重负。从启动（例如，通过社会措施）到效果（例如，在感染率的变化中）的延迟的存在以及启动机制的巨大不确定性构成了重大的技术挑战，必须通过强有力的决策政策来解决。这个项目还将为学生提供系统建模、识别和控制领域的培训机会。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The primary goal of the proposal is to develop a set of tools to inform and advise decision makers to control large-scale global pandemics. Specific research problems to be addressed include (1) Making reliable predictions for the state of the pandemic in the next few days/weeks/months. (2) Estimating the effect of governmental measures in the dynamics of the pandemic. (3) Designing protocols for effective control of a pandemic to prevent overloading the healthcare system. Efficient management of large-scale epidemics that are likely to overwhelm the health care system on a national scale, which would surely lead to a significant increase in the loss of human lives, including deaths not related to the specific pathogen causing the epidemic. This project will make contributions to the fundamental research on developing actionable models for epidemic estimation and control. Specifically, the focus is on models that are amenable to (i) the estimation of model parameters, (ii) the estimation of the state of the epidemic, and (iii) computing optimal intervention policies; all three with high degrees of confidence based on the relatively small datasets that are available while an epidemic is evolving. The research includes data-driven techniques to determine the impact of non-pharmaceutical social measures for epidemic management. A key challenge in this area is to reliably infer causal relationships between social measures and their effect in the epidemic dynamics. Methods will be developed to determine intervention policies that minimize the loss of human life and economic impact, without overwhelming the healthcare system. The existence of delays from actuation (e.g., through social measures) to effects (e.g., in changes of infection rates) and the large uncertainty in the actuation mechanisms pose significant technical challenges that must be addressed through robust decision policies. This project will also provide training opportunities for students in the areas of system modeling, identification, and control.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.