Prediction of COVID-19 Patients using Machine Learning Algorithms

Abstract

Coronavirus disease (COVID-19), also known as severe acute respiratory syndrome (SARS-COV2), has caused widespread concern for public health worldwide. Based on its rapid spread among those exposed to the wet animal market in Wuhan, China, the city was identified as its origin. The symptoms, reactions, and rates of recovery observed in coronavirus patients around the world have varied. The number of sufferers continues to grow at an exponential rate, and some countries are currently dealing with the third wave. Since the most effective treatment for this disease has yet to be established, early discovery of probable COVID-19 patients can help isolate them socially, slowing the spread and flattening the curve This study examines current research on coronavirus disease and its impact across age groups. We evaluate the effectiveness of Decision Tree (DT), and Logistic Regression (LR) in detecting COVID-19 in patients based on symptoms. A dataset from a public repository was pre-processed before applying Machine Learning (ML) techniques to it. The results show that all ML algorithms are effective in identifying COVID-19 in potential patients. DT classifiers have the highest accuracy of 98.70%, while SVM, KNN, and LR algorithms achieve 93.60%, 93.50%, and 92.80%, respectively.