Analysis and Prediction of COVID-19 using Machine Learning

Inorder to support the control and containment of the virus, accurate and efficient predictive models must be developed. The COVID-19 pandemic has presented hitherto unheard-of challenges for international health systems. This work offers a novel approach for COVID-19 infection prediction using binary classification models, specifically Random Forest and XGBoost algorithms. These models are well-known for their excellent accuracy and efficiency when handling complicated datasets. This research focuses on utilizing the advantages of ensemble learning techniques to improve prediction accuracy, in contrast to earlier approaches that depended on different machine learning techniques with differing degrees of success. The technique involves a comprehensive preprocessing of the data to generate a dataset containing COVID-19 symptoms and patient demographics. Next, the XGBoost and Random Forest models are used to classify individuals as COVID-19 either beneficial or detrimental. The aim of the research is to determine the predictive power of these models for COVID19 infections through a comparative analysis of the models under various experimental conditions.