“…Among them, the simplest is RSM, which is basically a polynomial regression. Independently of the adopted technique, the general form of the method results in an approximation of the exact solution, 𝑦 = 𝑦 + 𝜖 (47) where, for a given state 𝑖 of input variables, 𝑦 is the exact solution from the numerical model, 𝑦 is the approximate response surface and 𝜖 is the error between the exact solution and the approximation. The approximations 𝑦 are those that minimize the sum of squared errors (SSE), defined as,…”