zbMATH — the first resource for mathematics

A choice of criterion parameters in a linearization of regression models. (English) Zbl 0845.62047
There are many results obtained in the theory of nonlinear regression models; nevertheless much more and simpler inferences may be made in linear models. Thus it is of some importance to analyze situations where a nonlinear model can be substituted by a linear one. Several parameters (criterion parameters) have to be chosen and it is shown that some relations among these parameters and some natural restrictions on them exist.

62J02 General nonlinear regression
62F10 Point estimation
62J05 Linear regression; mixed models
Full Text: EMIS EuDML