COVID-19 Recovery Prediction Using Regression-Based Machine Learning Approaches

ABSTRACT

The number of people affected by the ongoing coronavirus has fluctuated rapidly and it has become strenuous to predict when will this pandemic end. To impede the spread of this virus, it is the need of the hour to maintain a social distance, wear masks and sanitize regularly. No doubt, the mortality rate has escalated, summing to a large percentage of population and destroying lives and economies with it. In India, mortality has climbed to as high as 3.2% as per the Indian Express, with recovery rate summing to 81.55% according to Times of India. Therefore, it has become prudent to determine and predict the effect and drawbacks of various factors such as testing, mortality rate and confirmed cases on the recovery rate. Due to progression and evolution in the discipline of machine learning, it has become practicable to get a middling figure of effects of these factors on death rate. Regression, one of the most broadly exhausted machine learning and statistics algorithm, is used to make predictions from data by learning the relationship between the features. In this article, regression algorithms are used to anticipate the same by using a cumulative data of all states in India. Study compares the ramification of the number of testing done and their impact on the recuperation of life due to the virus. Therefore, based on the research and computing, it was found that ridge regression gave the highest accuracy equivalent to 99.6%. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.