Table 1.
Case study 1. Mean-field and learned DE models for the BDM process for various values of (Pp, Pd). MSE denotes the mean squared error between the model solution and 〈CABM(t)〉.
| Pp | Pd | mean-field model (MSE) | learned model (MSE) |
|---|---|---|---|
| 0.01 | 0.005 | dC/dt = 0.005C − 0.01C2 (0.0011) | dC/dt = 0.00468C − 0.0095C2 − 0.0C3 (0.0001) |
| 0.05 | 0.025 | dC/dt = 0.025C − 0.05C2 (0.0026) | dC/dt = 0.02134C − 0.04572C2 + 0.00343C3 (0.0002) |
| 0.1 | 0.05 | dC/dt = 0.05C − 0.1C2 (0.004) | dC/dt = 0.03962C − 0.09057C2 + 0.01561C3 (0.0003) |
| 0.5 | 0.25 | dC/dt = 0.25C − 0.5C2 (0.01) | dC/dt = 0.15671C − 0.49984C2 + 0.33125C3 (0.0005) |