| 1 |
-
•
The date when a COVID-19 case was first reported in a region was set as the time instance t = 1.
-
•
The number of infected individuals in hospitals together with the accumulated numbers of cured and dead individuals during the period t to t + 5 were inserted into the SEIR model.
-
•
The initial number of susceptible individuals was equal to the total number of people in the region, and the initial number of exposed individuals was set according to the reported number of people in close contact with infected individuals.
|
| 2 |
-
•
The particle swarm optimization (PSO) technique was used to simulate the SEIR model and obtain the PSO-fitted values of β
1, γ and λ.
-
•
Given that these three parameters represent average transmission, cure and morality rates during the 6 day-window period, they were regarded as the real values during the period t + 1 to t + 2.
|
| 3 |
-
•
The parameters β
1, γ and λ were inserted into the SEIR model to calculate the possible numbers of susceptible and exposed individuals at time t + 2.
-
•
The moving window was forwarded to the period t + 1 to t + 6, and the step 2 was repeated to obtain the new fitted values of β
1, γ and λ.
|
| 4 |
-
•
After all the available data had been processed, the β
1, γ and λ values obtained by all moving windows were used to simulate the number of infected individuals during the time period of interest.
-
•
Goodness of fit was assessed in terms of the mean absolute error (MAE), the root mean squared error (RMSE) and the coefficient of determination (R
2) of the linear regression between the simulated and the actual numbers of infected individuals.
|
