Table 2.
Models used for forecasting recovered cases and their parameters.
| No. | Country | RNN model | Epochs | Hidden size | Number of layers | Learning rate | MSE | RMSE |
|---|---|---|---|---|---|---|---|---|
| 1 | USA | GRU | 2.50E+02 | 3.00E+02 | 2.00E+00 | 1.00E-05 | 3.61E+11 | 6.01E+05 |
| LSTM | 2.30E+01 | 3.00E+02 | 3.00E+00 | 1.00E-05 | 3.72E+11 | 6.10E+05 | ||
| 2 | Brazil | GRU | 6.56E+02 | 3.00E+02 | 2.00E+00 | 1.00E-05 | 8.53E+09 | 9.24E+04 |
| LSTM | 4.02E+02 | 3.00E+02 | 2.00E+00 | 1.00E-05 | 1.44E+12 | 1.20E+06 | ||
| 3 | India | GRU | 1.52E+03 | 3.00E+02 | 2.00E+00 | 1.00E-05 | 7.67E+04 | 2.76E+02 |
| LSTM | 1.26E+03 | 3.00E+02 | 2.00E+00 | 1.00E-05 | 6.62E+04 | 2.57E+02 | ||
| 4 | Russia | GRU | 3.00E+02 | 3.00E+02 | 2.00E+00 | 1.00E-05 | 1.17E+06 | 1.08E+03 |
| LSTM | 1.00E+02 | 3.00E+02 | 2.00E+00 | 1.00E-05 | 7.65E+06 | 2.77E+03 | ||
| 5 | South Africa | GRU | 2.70E+03 | 3.00E+02 | 2.00E+00 | 1.00E-05 | 1.67E+07 | 4.08E+03 |
| LSTM | 2.00E+03 | 3.00E+02 | 2.00E+00 | 1.00E-05 | 2.15E+06 | 4.05E+03 | ||
| 6 | Mexico | GRU | 6.50E+02 | 3.00E+02 | 2.00E+00 | 1.00E-05 | 1.61E+08 | 1.27E+04 |
| LSTM | 1.65E+03 | 3.00E+02 | 2.00E+00 | 1.00E-05 | 1.73E+08 | 1.32E+04 | ||
| 7 | Peru | GRU | 2.66E+02 | 3.00E+02 | 2.00E+00 | 1.00E-05 | 6.56E+06 | 2.56E+03 |
| LSTM | 7.50E+01 | 3.00E+02 | 2.00E+00 | 1.00E-05 | 2.13E+07 | 4.61E+03 | ||
| 8 | Chile | GRU | 3.00E+02 | 3.00E+02 | 2.00E+00 | 1.00E-05 | 1.32E+06 | 1.15E+03 |
| LSTM | 2.60E+02 | 3.00E+02 | 2.00E+00 | 1.00E-05 | 7.65E+05 | 8.74E+02 | ||
| 9 | UK | GRU | 3.00E+02 | 3.00E+02 | 2.00E+00 | 1.00E-05 | 1.19E+01 | 3.40E+00 |
| LSTM | 6.46E+02 | 3.00E+02 | 2.00E+00 | 1.00E-05 | 9.20E+00 | 3.03E+00 | ||
| 10 | Iran | GRU | 7.00E+02 | 3.00E+02 | 2.00E+00 | 1.00E-05 | 1.09E+06 | 1.04E+03 |