Application of Horizontal Federated Learning for Critical Resource Allocation - Lessons from the COVID-19 Pandemic

Abstract

The COVID-19 pandemic has caused disruption all over the industry. Significantly healthcare systems have been most affected due to a scarcity and acute demand for critical hospital resources. In this paper, we have proposed a framework for resource allocation using the characteristics of horizontal federated learning. We have used data of COVID cases in India and its states. We sourced the available critical hospital resources data and conducted the experimental analysis of the ten most vulnerable states of the country. The resource allocation method relies on the severity, spectrum of disease, and the patient’s length-of-stay in the hospital. The overall proposed methodology manages the limited resources and optimizes its use for the upcoming COVID cases. The collaborative feature of federated learning in the paper helped to update the information on patients, resources, and infection rates of different states, which in turn helped declare the overall severity of the pandemic in the country. Going further, our study will be helpful in the healthcare system’s authority to plan on hospital resources and manage them efficiently.