An Analysis Comparing the Predictions of Predictive Models with the Actual Number of New Cases during the COVID-19 Pandemic

Abstract

On January 3, 2020, Chinese health  officials discovered a pneumonia outbreak in the Chinese  city of Wuhan. The virus swiftly spread to most countries,  infecting a substantial portion of the population. On  September 28, 2020, about a million deaths were reported  worldwide. The virus is typically transmitted through  coughing or sneezing on others. Because there is no  effective vaccine, the majority of governments have  enforced lockdowns to slow the virus's spread. Other  preventive measures, such as travel bans, social isolation,  hand hygiene, and the use of face masks, were critical in  restricting the virus's spread. However, it proved difficult  to contain the virus, which had spread to over 200 nations  with a population of over 7 billion people. Large data sets  were acquired on a daily basis, and data analytics became  critical for uncovering trends and establishing how the  infection spread. Several studies have been conducted to  develop a mathematical forecast for the pandemic since the  disease's first cases in India. First examined with reference  to India, these models differed significantly in their scope,  underlying assumptions, and numerical forecasts. The  objective of this research is to evaluate the predictive  models' efficacy by comparing their forecasts with the  actual number of new cases reported every day in India  during the COVID-19 outbreak.