Previous |  Up |  Next


nonlinear regression model; parameter-effects arrays; confidence region; three-dimensional parameter effects; Bates-Watts array; four-dimensional parameter-effects arrays; transformations of parameters
Formulas for a new three- and four-dimensional parameter-effects arrays corresponding to transformations of parameters in non-linear regression models are given. These formulae make the construction of the confidence regions for parameters easier. An example is presented which shows that some care is necessary when a new array is computed.
[1] Bates D. M., Watts D. G.: Relative curvature measure of nonlinearity. Journal of the Royal Statistical Society (1980), 1-25. MR 0567196
[2] Bates D. M., Watts D. G.: Parameter transformations for improved approximate confidence regions in nonlinear least squares. The Annals of Statistics 9 (1981), 1152-1162. DOI 10.1214/aos/1176345633 | MR 0630099 | Zbl 0493.62060
[3] Clarke G. P. I.: Marginal curvatures and their usefulness in the analysis of nonlinear regression models. Journal of the American Statistical Association 82 (1987), 844-850. DOI 10.1080/01621459.1987.10478507 | MR 0909991 | Zbl 0623.62062
[4] Cook R. D., Goldberg M. L.: Curvatures for parameter subsets in nonlinear regression. The Annals of Statistics 14 (1986), 1399-1418. DOI 10.1214/aos/1176350166 | MR 0868308 | Zbl 0651.62060
[5] Cook R. D., Witmer J. A.: J. Amer. Statist. Assoc. 80 (1985), 872-878. MR 0819586
[6] Hoogaard P.: Parameter transformations in multiparameter nonlinear regression models. University of Copenhagen, Institute of Mathematical Statistics, 1984, Preprint No. 2.
[7] Potocký R., Van Ban T.: Confidence regions in non-linear regression models. Applications of Mathematics 57 (1992), 29-39.
Partner of
EuDML logo