Prediction of COVID-19 using machine learning techniques

Abstract

In December 2019, the world faced an unprecedented and formidable challenge in the form of COVID-19. Since its first reported case in December 2019 and subsequent classification as a pandemic by the World Health Organization (WHO), it has imposed an enormous impact, taking lives, disrupting diverse economic sectors, and introducing numerous challenges. Predicting and controlling COVID-19 precisely remains a pivotal concern for the future. To enhance the precision of COVID-19 disease prediction and alleviate the burden on healthcare systems, we explore the application of diverse machine learning classifiers. Leveraging datasets comprising confirmed cases, recoveries, and fatalities, our study seeks to improve the predictive accuracy, ensuring efficient and precise evaluation and triage of patients. This, in turn, the load on overburdened emergency departments is reduced, and mitigates the pressures faced by healthcare professionals. The pivotal role of artificial intelligence (AI) and machine learning (ML) in this context cannot be overstated. Our research endeavors to refine the accuracy and quality of results by employing advanced machine learning techniques such as Random Forest (RF), Decision Tree (DT), Support Vector Machine (SVM), Naïve Bayes (NB), Gradient Boosting (GB), and Extreme Gradient Boosting (XGBoost). Our dataset is sourced from Kaggle, a well-established platform for data science and machine learning. Our comprehensive analysis compares the outcomes of six distinct models. Notably, the Gradient Boosting classifier surpasses other techniques, achieving an impressive accuracy rate of 90% for confirmed cases, 90% for recoveries, and 92% for fatalities. This represents a significant improvement over the baseline paper, which achieved accuracy rates of 83% for confirmed cases, 72% for recoveries, and 81% for fatalities using the same dataset. Further our research enhances COVID-19 prediction, aiding healthcare professionals in addressing the other global epidemic for future.