Answers are grounded in this paper and cite their sources.
The first step to an analysis of experimental data is commonly curve fitting; to obtain meaningfnl parameters, data are graduated by a theoretically established functional form or by an empirical expression. In either case, the parameters most often appear in a nonlinear form in the mathematical expression that describes the data. Least-squares and maximum likelihood graduating methods are difficult, if not impossible, to apply to many expressions. The least-squares criterion for graduating is refuted by nonlinear parameters. To use least-squares with nonlinear parameters, it is necessary to restrict the parameters and to use iterative computer routines.',* We propose that a graduating rriterion to conserve the energy of a system by conserving the first and higher moments is more pertinent than minimizing the linear deviations from a curve. For example, the area under a stress-strain curve defined by data should equal that defined by a graduated relationship. We have developed mathematics and graduating techniques to cope with nonlinear parameters and to conserve the energy of a system. The technique is applicable to many fields of analysis, but this particular effort was directed to polymer physical properties data.
No takes yet. Share an insight, caveat, or question.
Lepper et al. (1968) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: