Table 6.
Case study 5: mean-field and learned DE models for the SIR ABM for various values of (PI, PR), along with the computed MSE values between model and ABM output, and model R0 calculations. We fix Pm = 1 in each scenario.
| PI | PR | mean-field model (MSE) | R0 | learned model (MSE) | R0 |
|---|---|---|---|---|---|
| 0.005 | 0.0005 | dS/dt = −0.0025IS (0.0027) | 5.0 | dS/dt = −0.00229IS (0.0012) | 4.68 |
| dI/dt = 0.0025IS − 0.0005I (0.002) | dI/dt = −0.00049I + 0.00225IS (0.0009) | ||||
| 0.01 | 0.001 | dS/dt = −0.005IS (0.0044) | 5.0 | dS/dt = −0.00447IS (0.0006) | 4.52 |
| dI/dt = 0.005IS − 0.001I (0.0029) | dI/dt = −0.00098I + 0.00443IS (0.0005) | ||||
| 0.05 | 0.005 | dS/dt = −0.025IS (0.0107) | 5.0 | dS/dt = −0.01881IS (0.003) | 3.72 |
| dI/dt = 0.025IS − 0.005I (0.0066) | dI/dt = −0.00472I + 0.01833IS (0.0021) | ||||
| 0.1 | 0.01 | dS/dt = −0.05IS (0.0164) | 5.0 | dS/dt = 0.01569S − 0.01608S2 − 0.00545I − 0.0459IS (0.0055) | 3.41 |
| dI/dt = 0.05IS − 0.01I (0.0097) | dI/dt = −0.00906I + 0.03101IS (0.0033) |