Exploring the role of country social and medical characteristics in patient level mortality in COVID-19 pandemic using Unsupervised Learning

Abstract

This work aims to contribute to the field of COVID-19 pandemic analysis. In this research we applied a twofold analysis that focused initially on the country general social-economic and medical characteristics and on top of that in a second level exploring the correlations to the characteristics that affect COVID-19 patients’ mortality level. The approach has been applied to large datasets that include country level medical and the socio-economic data according to World Health Organization, the role of the cigarette consumption per capita using open datasets, and the cumulative data of the “COVID-19 Data Repository by the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University” for the total number of Cases, Deaths and Recovered. 101 countries including twenty-two (22) features are studied. We have also drilled in the country of Mexico datasets to show case the effectiveness of our approach. We show that our approach can achieve 96% overall accuracy based on the proposed combination approach of macro and micro features. Our approach outdoes previous study results that utilize machine learning to assist medical decision-making in COVID-19 prognosis. We conclude that country social economic and medical characteristics play important role to COVID-19 patients’ prognosis and their outcome.