Modeling and Prediction of COVID-19 in India Using Machine Learning

Ghosh, A.; Sadhu, A.

The inculcation of efficient forecasting and prediction models may assist the government in implementing better design strategies to prevent the spread of the virus. However, most of the machine learning methods along with SEIR models have failed to predict the outcome valid for a longer duration. In this paper, we propose a simple yet effective method for modeling COVID19 like pandemic based on a non-linear regression curve. With the help of machine learning tools like recurrent neural networks and artificial neural networks, we predict the values of the regression coefficient and regression power. We also evaluate the effectiveness of different measurements taken to combat the pandemic situation in India. The proposed method outperforms the existing techniques for prediction. This, alongside giving an overview and prediction of the COVID19 in India, would help to model future pandemics. © 2021, Springer Nature Switzerland AG.