×

Uniqueness of surfaces and global identifiability in nonlinear regression models. (English) Zbl 0625.62044

The parameter of a nonlinear regression model is determined uniquely for almost all experimental designs if the number of experimental points is greater than the dimension of the parameter and different regression functions have no weak contact of infinite order. This result yields global identifiability for most practically used curve-fitting models, especially with polynomial, exponential and trigonometric regression functions. In this approach regression models with errors of measurement in the independent variables are included as well.
The main theorem does not belong to statistics. It establishes the geometric fact that parameter-dependent \(C^{\infty}\)-surfaces are almost surely determined uniquely by sufficiently many base points. The proof shows that the singular zeros of sparse \(C^{\infty}\)-mappings can be imbedded in a countable union of \(C^{\infty}\)-manifolds with a sufficiently small dimension provided the coordinate functions have no weak contact of infinite order for different parameters.

MSC:

62J02 General nonlinear regression
57N75 General position and transversality
93B30 System identification
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] DOI: 10.2307/1909759 · Zbl 0384.62090 · doi:10.2307/1909759
[2] Dieudoniste J., Foundations of modern analysis (1960)
[3] Diehdoistne J., Grundzuge der modemen Analysis (1976)
[4] Fisher F.M., The identification problem in econometrics (1966)
[5] Fuller W. A., Ann. Statist 1 pp 539– (1982)
[6] DOI: 10.1007/978-1-4615-7904-5 · doi:10.1007/978-1-4615-7904-5
[7] Hoschel H.P., Transactions of the S-th Prague Conference on Information Theory, Statistical Decision Functions and Random Processes pp 305– (1978)
[8] Hoschel H.P., Miihlhausen G.D.R pp 55– (1979)
[9] Hoschel H.P., Least-squares fitting for nonlinear models with errors in variables and globally convergent Gauss-Newton-procedures. Discussion paper 8025, Center for Operations Research and JKcoiioineiries (1980)
[10] Hoschiil H.P., Modelle mit Fehlern In den Variablen. In HUMAK, K.M.S. Statistische Methoden der modellbildung (1983)
[11] Humak K.M.S., statistische Methoden der Modellbildung (1977) · Zbl 0353.62001
[12] DOI: 10.1214/aoms/1177697731 · Zbl 0193.47201 · doi:10.1214/aoms/1177697731
[13] Karlin S., Tchebycheff systeme with applications in analysis and statistics (1966) · Zbl 0153.38902
[14] Pazman A., Math Operutionsforsch u Statist 15 pp 323– (1984)
[15] DOI: 10.2307/1913267 · Zbl 0231.62081 · doi:10.2307/1913267
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.