Predictions for COVID-19 outbreak in India using Epidemiological models

Rajesh Ranjan

Results for the SIR model, which has better abilities for long term forecasts, are shown in Figure 3 . The statistical parameters for this regression model is also included in Table 2 , with a coefficient of determination of R 2 = 0.992, and p-value very close to zero indicating high statistical significance. As seen in the top panel of Figure 3 , after the initial exponential phase till March 31, the acceleration phase starts and continues till April 13 when there is a peak of around 400 cases per day (bottom panel). The deceleration phase begins from this date and continues till April 30 after which asymptotic slow growth begins till flattening of the curve. The estimates from this model reaches a plateau at 12,416 cases, which is the final epidemic size. In the acceleration phase (April 1-April 13), the predictions from the SIR model differ from exponential model. For example, the exponential model suggests that the number of cases will cross 5,000 on April 6, but the prediction for the same using SIR model is on 12 April.