Global Prediction of COVID-19 Cases and Deaths using Machine Learning

Abstract

Coronavirus Disease or COVID-19 pandemic has taken over the world by storm. It has horrifying effect on the health of the people. Continuously rising number of COVID-19 cases has and still creating huge stress on the governing bodies of all countries, and they are finding it hard to find solution for the situation. This project's goal is to explore machine learning and develop a COVID-19 model that can predict number of cases with high accuracy. The proposed study employs SVR and PR models to forecast the number of recovered cases, confirmed cases, deaths, and daily case count. The data is collected from the 1st of March to the 30th of April 2020. The confirmed number of cases as of April 30th were 35043, with 1147 total deaths and 8889 recovered patients. The model was created in Python 3.8.5. We will look at various machine learning prediction algorithms and compare them. In conclusion, supervised learning algorithms proved to be better than unsupervised learning algorithms. These prediction models can help us to brace for another COVID-19 wave and to ensure the availability of the required resources.