Analysis and Prognosis of COVID-19 Pandemic in India - A Machine Learning Approach

Abstract

Machine Learning techniques have been used for decades to analyze and predict outcomes in various domains of study as well as in industries. It employs various algorithms to perform different tasks. One of the practical and topical applications in the current scenario is to use the power of Machine Learning to study various aspects of the ongoing pandemic (COVID-19), since the entire world is in its grip. COVID-19 was declared a global pandemic on 11 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">th</sup> March 2020 by WHO. Worldwide more than 43 Million people have contracted this viral disease and more than 1.1 million people have succumbed to it (as on 27th October 2020). The number of people affected by COVID-19 in India is increasing at a fast pace and currently India has the second highest number of cases and third highest casualties in the world. In this study, a model is developed to examine and analyze the spread of this disease. Four different Machine Learning algorithms namely Random Forest Regression, Multiple Linear Regression, Support Vector Regression and Lasso Regression have been considered. A Kaggle dataset consisting of figures of confirmed cases, patients recovered, and people that have died due to COVID-19 across India over a particular period of time has been used. The results of this study indicate that Random Forest Regression provides the most accurate results whereas Support Vector Regression is least accurate.