Simple Forecasting Model for COVID-19 Cases in India - Multilevel Model Evaluation with R2, MSE, and MAE
European Journal of Molecular & Clinical Medicine,
2020, Volume 7, Issue 11, Pages 2130-2142
AbstractAll-inclusive, more than 20 million individuals have been infected with the COVID-19, with the most number of cases from the United States, trailed by Brazil and India. In perspective on this developing extent of the number of cases, forecasting models are exceptionally useful to be prepared to confront the pandemic circumstance. In this work, we have used an efficient time series based Machine Learning (ML) algorithm to forecast the COVID-19 cases in India. We have trained the system with data from 3 March 2020 to 7 August 2020 and we have forecasted the values from 8 August 2020 to 9 September 2020. We have seen that the total no. of cases will get doubled, i.e. reaches 40 Lakhs by the end of the forecasted period. Along with this forecast we have done the multi-level validation of our work using metrics, r-squared error (R2), mean squared error (MSE), mean absolute error (MAE).
- Article View: 231
- PDF Download: 284