Figure 7

In the physical constraints design pattern, the computation in the data-driven block \(\mathcal{D}\) is informed by the domain knowledge in \(\mathcal{P}\). The constraints can affect the architecture of \(\mathcal{D}\), its parameters, or computational results both at intermediate levels and at the output. We distinguish between hard constraints (Fig. 7a) which take effect both during the model fitting stage and inference time and soft constraints (Fig. 7b) which are typically only applied at training time. After training soft constraints are implicitly encoded in \(\mathcal{D}\), but no longer used explicitly. For this reason, we denote the soft constraints in a dashed manner