Skip to Main Content
The author discusses some of the problems associated with data fitting, particularly the case of uncertainties in all the variables, the possibly misleading results of algorithmic convergence, and also the limitations imposed by some commercial software. He suggests that the presence of local minima in the fitting function means that the parameters derived from the fitting may be subject to much larger uncertainty than often is suggested. He also addresses the difficulty of making accurate numerical Laplace transforms, even when regularization techniques are employed. His general conclusion is that the parameters derived by modeling dielectric data are liable to have considerable uncertainty.