<![CDATA[The severity of COVID-19 varies in each country, requiring an analytical approach that can provide accurate classification as a basis for global health decision-making. Machine learning methods are an effective option for detailing the severity based on data patterns regarding cases, deaths, and other indicators. In this study, the K-Nearest Neighbor (KNN) algorithm was compared with Support Vector Machine (SVM) using a global dataset on COVID-19 taken from Kaggle. The analysis process included data pre-processing, data exploration, model building, and evaluation using accuracy, precision, recall, and F1-score metrics. The results of the evaluation showed that SVM performed better with an accuracy of 87%, while KNN only reached 78%. In addition, SVM also produced a lower and more consistent classification error rate in each severity category. Based on these findings, SVM is considered more efficient in classifying the severity of COVID-19 in globally distributed data that is unevenly distributed.]]>
Copyrights © 2025
