Information criteria for the choice of parametric functions for measurement

This paper deals with the problem of nonlinear reconstruction of a measurand or the so-called nonlinear inverse problem. Some criteria leading to the choice of the parametric models to be used in inverse problems are presented. It is demonstrated that some parametric functions are “better” than others when the final goal is an accurate estimate of some unknown quantity. Examples dealing with analytical functions or with a more realistic heat conduction problem are given