Optimization with different parameters under lockdown and unlockdown periods for forecasting and prediction for covid-19

There are many infectious diseases which result in a pandemic in the countries they affect The currently spreading world-wide pandemic is COVID-19, which is caused by the coronavirus which causes respiratory issues and serious illness to the elderly people It is mainly transmitted from the nasal discharge when the infected person coughs or sneezes In this paper, prediction and forecasting of the trajectory of spread of this infectious disease is done using machine learning methods like linear regression algorithm and XGBoost algorithm A time series prediction is also made using Prophet By these methods, a comparison is drawn between India and USA from January 22, 2020 till July 31, 2020 for the cases like confirmed infection, people who died due to this virus and those who recovered from the virus, during the lockdown periods as well as in the un-lockdown periods in India, along with other parameters such as Temperature (high/ low) in degree Celsius and Reproduction number (R0) for each day, whether BCG vaccine is administered or not and population of each country It is found that even though USA has a lesser population compared to India, and the days under complete lockdown were more, the cases in USA reach the highest in confirmed infection, people who succumbed to the infection and those who recovered The forecasting is done for 90 days and the prediction shows that for upcoming days, the cases for confirmed infection, people who died and those who recovered will be more or less similar to that of the previous month and slight decrease in the count of people in the above 3 categories © 2021, Annals of the Romanian Society for Cell Biology All rights reserved