Abstract:
Corona viruses (CoV) are part of a family of viruses that cause illness ranging from the
flu to more severe diseases such as Middle East Respiratory Syndrome (MERS-CoV) and
Severe Acute Respiratory Syndrome (SARS-CoV). The disease caused by the corona virus,
also known as COVID-19. Many research have been done to build a clustering system
especially on health department, in particular for COVID-19 study. This research aimed
to cluster COVID-19 patients so that patients get the right treatment, it can also be an
illustration for the medical side of COVID-19 patients treated from April to June 2021 at
the Arifin Achmad Hospital, Riau Province and measure the level of accuracy of the
method used. This study used the K-Means Clustering method in grouping data and used
the Silhouette Coefficient to see how good the quality of the resulting cluster is. The stages
in this research are data preparation, data preprocessing, clustering process with KMeans,
testing clustering results using Silhouette Coefficient and concluding the results.
Based on the results of the study, that grouped the status of COVID-19 patients used 20
variables with the number of clusters K = 3 has a Silhouette Coefficient value of
0.15653384 included the no structure group, from the 3 clusters owned, it can be seen
that the highest s(i) is in the cluster that is close to number 1, namely cluster 1 with a
value of 0.48356834. So clustered the COVID-19 patient data with a total of 500 patients
used the K-Means Clustering method did not achieve satisfactory results, because there
were still many cluster members who did not fit into the proper cluster group. One
example of cluster members who do not fit into the proper cluster group can be seen in
cluster 1, where there are still patients with mild and moderate symptoms, as well as in
other clusters.