1 Parametric methods
These methods are premised on the principle that the changes in the response variable(s) of interest (e.g. blood pressure) over time within a complicated model can be approximated by a reduced-order model [30]. To illustrate this concept, consider the 3-dimensional relationship illustrated in figure 3A. If we look at this system from a 2-dimensional X-Z perspective (figure 3B), we see that Z only varies within a relatively narrow range for any given value of X. Therefore Z may be approximated from a known value of X regardless of the value of Y. Parametric model-order reduction methods attempt to achieve the same goal but differ in the way they find the simplified model.
Parametric model-order reduction methods can be divided into two main categories; (i) methods which focus on reducing the number of nodes (e.g. state variables or compartments) in the model, and (ii) methods that focus on reducing the number of edges (i.e. reactions and interactions). The following section will provide an overview of both approaches.