Zur Quantfizierung und Analyse der Nichtlinearität von Regressionsmodellen
DOI:
https://doi.org/10.17713/ajs.v27i3.537Abstract
Bei nichtlinearen Regressionsmodellen treffen im endlichen Stichprobenfall eine Reihe von statistischen Aussagen nicht zu, die für das lineare Regressionsmodell gelten. So ist die (iterativ berechnete) Kleinste-Quadrate-Summen-Schätzfunktion (KQS-Schätzfunktion) imnichtlinearen Fall nicht effizient, wenn sie auch unter gewissen Voraussetzungen asymptotisch effizient ist. Für nichtlineare Regressionsmodelle werden daher statistische Aussagen gerne mit Hilfe der asymptotischen Theorie getroffen. Zeigt das Regressionsmodell aber ausgeprägt nichtlineares Verhalten, so führen so erhaltene Aussagen zu falschen Ergebnissen. Nichtlinearitätsmaße wie die von Bates and Watts (1980) geben Aufschluß darüber, wie ausgeprägt die Nichtlinearität eines Regressions-modells ist. Passende Parametertransformationen sowie die Verwendung eines geeigneten Versuchsplans sind Möglichkeiten, die Nichtlinearität zu reduzieren, sodaß mit Hilfe von linearen Approximationen getroffene Aussagen zu validen Ergebnissen führen. Eine Fallstudie illustriert die Auswirkungen der Wahl der Versuchspunkte auf die Nichtlinearität eines Modells; die Verwendung eines geeigneten Versuchsplans kann zu einer
deutlichen Reduzierung der Nichtlinearität führen.
References
D.M. Bates and D.G. Watts. Relative curvature measures of nonlinearity (with discussion). Journal of the Royal Statistical Society, Ser. B, 42:1-25, 1980.
D.M. Bates and D.G. Watts. Parameter transformations for improved approximate confidence regions in nonlinear least squares. Annals of Statistics, 9:1152-1167, 1981.
D.M. Bates and D.G. Watts. Nonlinear Regression Analysis and its Applications. Wiley & Sons, New York, 1988.
E.M.L. Beale. Confidence regions in non-linear estimation (with discussion). Journal of the Royal Statistical Society, Ser. B, 22:41-88, 1960.
G.E.P. Box and H.L. Lucas. Design of experiments in nonlinear situations. Biometrika, 47:77-99, 1959.
M.J. Box. Bias in nonlinear estimation (with discussion). Journal of the Royal Statistical Society, Ser. B, 33:171-201, 1971.
G.P.Y. Clarke. Moments of the least squares estimators in a non-linear regression model. Journal of the Royal Statistical Society, Ser. B, 42:227-237, 1980.
G.P.Y. Clarke and L.M. Haines. Optimal design for models incorporating the richards function. In G.U.H. Seeber, B.J. Francis, R. Hatzinger, and G. Steckel-Berger, editors, Statistical Modelling - Proceedings of the 10th International Workshop on Statistical Modelling, pages 61-66, 1995.
M.A. Clyde. Bayesian design for approximate normality. In C.P. Kitsos and W.G. Müller, editors, MODA 4 - Advances in Model-Oriented Data Analysis, pages 25-37, 1995.
D.C. Hamilton and D.G. Watts. A quadratic design criterion for precise estimation in nonlinear regression models. Technometrics, 27:241-250, 1985.
A. Pázman. Nonlinear Statistical Models. Kluwer Academic Publishers, Dordrecht/Boston/London, 1993.
A. Pázman and L. Pronzato. Nonlinear experimental design based on the distribution of estimators. Journal of Statistical Planning and Inference, 33:385-403, 1992.
D. Rasch. The robustness against parameter variation of exact locally optimum designs in nonlinear regression - a case study. Computational Statistics and Data Analysis, 20:441-453, 1995.
D.A. Ratkowsky. Nonlinear Regression Modeling. Marcel Dekker, New York, 1983.
G.A.F. Seber and C.J. Wild. Nonlinear Regression. Wiley & Sons, New York, 1989.
G. Sedlacek. Nichtlineare Regression und Versuchsplanung. Diplomarbeit, Technische Universität Wien, 1996.
W.N. Venables and B.D. Ripley. Modern Applied Statistics with S-PLUS. Springer, New York, 1997.
Downloads
Published
Issue
Section
License
The Austrian Journal of Statistics publish open access articles under the terms of the Creative Commons Attribution (CC BY) License.
The Creative Commons Attribution License (CC-BY) allows users to copy, distribute and transmit an article, adapt the article and make commercial use of the article. The CC BY license permits commercial and non-commercial re-use of an open access article, as long as the author is properly attributed.
Copyright on any research article published by the Austrian Journal of Statistics is retained by the author(s). Authors grant the Austrian Journal of Statistics a license to publish the article and identify itself as the original publisher. Authors also grant any third party the right to use the article freely as long as its original authors, citation details and publisher are identified.
Manuscripts should be unpublished and not be under consideration for publication elsewhere. By submitting an article, the author(s) certify that the article is their original work, that they have the right to submit the article for publication, and that they can grant the above license.