Abstract
A T k criterion is proposed and investigated, which makes it possible to compare two competing models when restoring a dependence from a single series of measurements. A method is developed for detecting false minima on the graph of the realization of the residual variance. A proof is given of the efficiency of utilizing the T k criterion to resolve questions of the significance of a parameter included in a model. An asymptotic T k criterion is obtained, considered, and recommended for introduction into a physical experiment.
Similar content being viewed by others
References
N. R. Draper and H. Smith, Applied Regression Analysis, Wiley, New York (1981).
G. A. F. Seber, Linear Regression Analysis, Wiley, New York (1977).
C. R. Rao, Linear Statistical Inference and its Application (2nd ed.), Wiley, New York (1973).
G. A. Ivanov and I. A. Krivosheev, Statistical Methods of Processing Experimental Data when Restoring a Dependence [in Russian], Dal'nauka, Vladivostok (1998).
G. A. Ivanov and A. F. Turbin, Statistical methods of Restoring a Dependence from Test Data [in Russian], Znanie, Kiev (1986).
G. A. Ivanov and Yu. R. Chashkin, Some Topics in the Theory of Random Processes [in Russian], Mathematics Institute, Academy of Sciences of the Ukrainian SSR, Kiev (1984), p. 133.
K. Mardia and P. Zemroch, Tables of F Distributions and Associated Distributions [in Russian], Moscow (1984).
S. Wilks, Mathematical Statistics [Russian translation], Nauka, Moscow (1967).
N. P. Klepikov and S. N. Sokolov, Analysis and Planning of Experiments by the Maximum Likelihood Method [in Russian], Nauka, Moscow (1964).
G. A. Ivanov, I. A. Krivosheev, and Yu. R. Chashkin, Izmerit. Tekh., No. 7, 8 (2000).
G. Ivanov and I. Krivosheyev, Proc. Pacific Intern. Conf. on Mathematical Models and Cryptography, Vladivostok (1995), p. 47.
Rights and permissions
About this article
Cite this article
Ivanov, G.A., Krivosheev, I.A. A Criterion for Comparing Models Approximating Experimental Data and Its Asymptotic Properties. Measurement Techniques 44, 785–794 (2001). https://doi.org/10.1023/A:1012936015681
Issue Date:
DOI: https://doi.org/10.1023/A:1012936015681