A Comparison of the COVID-19 Risk Prediction Model with Data Mining Techniques

Abstract

The purpose of this study was to compare the effectiveness of models in predicting the likelihood of contracting COVID-19. Three data mining techniques were used, namely, decision tree method Naive Bayes method and K-Nearest Neighbor method The efficacy of appropriate identification classification models for predicting the likelihood of contracting COVID-19 was compared using 5,434 rows of 21 columns of COVID-19 data from the kaggle website. Data were analyzed on the basis of CRISP-DM method. Use RapidMiner. using Rapidminer. in modeling in the analysis of data, accuracy, precision, recall. The results showed that the decision tree method was the most effective. The accuracy was 97.85%, followed by the K-Nearest Neighbor method, which gave an accuracy of 97.36%, and Naive Bayes method gave an accuracy of 97.36%. 96.75% from the results of this performance comparison. can adopt the decision tree method can be used as a model for predicting the risk of contracting COVID-19.