Hybridized wavelet neuronal learning-based modelling to predict novel COVID-19 effects in India and USA

Abstract

Coronavirus disease so called as COVID-19 is an infectious disease and its spread takes place due to human interaction by their pathogen materials during coughing and sneezing. COVID-19 is basically a respiratory disease as evidence proved that a large number of infected people died due to short breathing. Most widely and uncontrollably spreading unknown viral genome infecting people worldwide was announced to be 2019–2020 nCoV by WHO on January 30, 2020. Based on the seriousness of its spread and unavailability of vaccination or any form of treatment, it was an immediate health-emergency of concern of international-level. The paper analyses effects of this virus in countries, such as India and United States on day-to-day basis because of their greater variability. In this study, various performance measures, such as root mean square error (RMSE), mean absolute error (MAE), coefficient of determination (R2) , mean absolute standard error (MASE) and mean absolute percentage error (MAPE) which characterize models’ performances. R2 value has been achieved to be closest to 1, i.e., 0.999 from Wavelet Neuronal Network Fuzzified Inferences’ Layered Multivariate Adaptive Regression Spline (WNNFIL–MARS) for both the countries’ data. It is important to capture the essence of this pandemic affecting millions of the population daily ever since its spread began from January, 2020.