|
AUTHOR(S): Uneb Gazder
|
TITLE Identifying Global Patterns for COVID-19 using Cluster Analysis |
 PDF |
ABSTRACT Clustering is an effective technique for unsupervised classification of data. K-means clustering is considered one of the most effective approaches for partitioning the data. In this study, k-means clustering was applied using squared Euclidean distance metric for identifying temporal patterns in the global data of infected and death cases of COVID-19. It was found that there was a significant shift in infected cases on 13, 20 and 27 March 2020, the last being for decrease in cases. The increase points could be attributed to increase testing and surveillance and spread of viruses across borders. On the other hand, the decrease in infected cases could be attributed to the closure of schools, businesses, sporting events and travel activities. Death cases mostly follow the increasing trend shown by infected cases with an approximate lag of 7 days. However, a breakpoint with increase in death cases was observed on 4 April 2020 which could be attributed to falsified medicines and equipment. A reduction in death cases is observed recently with possible explanation being increased knowledge for treating the infected persons and managing the health care facilities. |
KEYWORDS unsupervised learning; infectious disease; COVID-19; clusters; temporal patterns |
|
Cite this paper Uneb Gazder. (2025) Identifying Global Patterns for COVID-19 using Cluster Analysis. International Journal of Mathematical and Computational Methods, 10, 180-186 |
|
