Statewise prediction of Covid-19 cases using machine learning, SEIR, and a time series model

Abstract

For anticipating post-operative outcomes, machine learning-based forecasting mechanisms have proven useful, which improves future decision-making. Machine Learning models have been used in numerous applications that require the identification of risk-related negative factors. Problems with forecasting are dealt with using a variety of prediction techniques. This paper shows the capacity of AI models to gauge the quantity of patients who'll get influenced by the Coronavirus, which is at present dynamic considered as a hazardous danger to humanity. Three outcomes are predicted by each model: the number of newly infected cases, fatalities, and recoveries. One important factor is to examine the manner in which the disease spreads state-by-state. to examine information like the number of infected individuals in each state, the number of infections in that state, and so on. We believe that state-wise forecasts would assist the federal government in better allocating its limited healthcare resources. The examination of spreading of COVID19 infection predictsthe size of the pandemic, alongside the recuperation rate and fatality rate.