Assessment of first-phase COVID-19 pandemic in Europe using hierarchical clustering based on principal components analysis

Document Type


Publication Date


Publication Title

Journal of Autonomous Intelligence


It is of great interest for researchers to assess the COVID-19 pandemic in Europe. Grouping of COVID-19-affected regions is an effective way to monitor and optimize planning to combat the disease. This paper applied hierarchical clustering based on principal components analysis (HCPCA) to COVID-19 data from affected European countries. Considering several attribute indices, we obtained a new set of indicators using principal components analysis to aggregate and reduce the dimension of attribute indices of affected countries. Further, we obtained groups of affected countries subject to their similarity using hierarchical clustering to the reduced observations of new attributes indices of these countries. This study aims to group European countries with similar epidemic severity using some presumed attribute indices. The study is limited up to 24 May 2020, to assess if the outputs of the study could help governments, administrators, World Health Organization (WHO), healthcare service professionals, and other decision-makers to optimize their policies and plan their regulations in the country level requirements so that transmission of infections, deaths, critical conditions of patients could be minimized. For this purpose, we used hierarchical clustering using principal components analysis to obtain better clusters of countries with similar epidemic severity.