COVID-19: Prediction of Confirmed Cases, Active Cases and Health Infrastructure Requirements for India

Abstract

To develop a holistic model for COVID-19 forecasting to determine the hospitalization needed in each Indian state. The “covid19 in India dataset from kaggle has been used to study the pattern of infection spread and to forecast the number of cases of covid19 patients in coming months. The dataset has different excel files which include the data of the „‟Confirmed‟, „‟Cured‟ and “Death” count in each state of India. Also, the dataset includes the number of beds in each state which is helpful to compare the current active cases vs. number of beds. The two models which have been used for forecasting the COVID-19 are “Prophet” and “ARIMA”. The outcome of this forecasting is then have been used to predict the number of beds would require providing the treatment to the infected people. The Prophet model has been implemented to forecast the number of beds required in high density states to accommodate the rising number of patient’s treatment. The study shows us that the number of cases would be exponential in coming months in absence of the vaccine and the preventive measures such as lockdown, social distancing, etc.

The model gives us the prediction that the COVID cases are rising exponentially. The number of predicted cases of infected people could help the government and public healthcare facilities to increase their capacity of healthcare services to provide the best care to the infected people.