Descriptions of nonlinear systems often involve a set of multivariate real functions. While very flexible, this has the downside of having a large number of parameters which makes gaining insight based on such models very difficult. When decoupling a set of multivariate real functions, these functions are expressed as a linear combination of univariate functions - removing cross terms. Such a decoupled representation may give more insight into the problem, reduce the number of variables or may be useful to simplify a complex system.