Abstract
Metamodels have been used frequently by the simulation community. However, not much research has been done with nonparametric metamodels compared with parametric metamodels. In this paper, smoothing splines for performing nonparametric metamodeling are presented. The use of smoothing splines on metamodeling fitting may provide functions that better approximate the behavior of the target simulation model, compared with linear and nonlinear regression metamodels. The smoothing splines tolerance parameter can be used to tune the smoothness of the resulting metamodel. A good experimental design is crucial for obtaining a better smoothing spline metamodel fitting, as illustrated in the examples.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
