Using Machine Learning to Estimate Unobserved COVID-19 Infections in North America

Detection of COVID-19 cases remains a huge challenge. The COVID-19 pandemic continues to take its toll; close to 2 million people are infected, and over 121,000 are dead. Dire projections are surfacing almost every day, and policymakers worldwide are using projections for critical decisions. Given this background, we model unobserved infections in North America to examine the extent to which we might be grossly underestimating COVID-19 infections in North America. We developed a machine learning model to uncover hidden patterns based on reported cases and to predict potential infections. First, our model relied on dimensionality reduction to identify key parameters that were key to uncovering hidden patterns. Next, our predictive analysis used an unbiased estimator approach to infer past infections from current fatalities.