Abstract
The article investigates the most common particular case of a confluent regression model for a passive experiment in which the variance is different for different observation vectors and homoskedastic within each observation vector. The regression parameters are estimated by the least distance method. The results are illustrated with an econometric example.
Similar content being viewed by others
References
A. S. Mechenov, “Partially approximate systems of linear algebraic equations,”Zh. Vych. Mat. Mat. Fiz.,31, No. 6, 720–799 (1991).
A. S. Mechenov, “Maximum likelihood approach to parameter estimation of linear functional relationships,” in:Numerical Methods in Mathematical Physics [in Russian], MGU, Moscow (1996), pp. 153–159.
A. S. Mechenov, “Confluent approach in regression analysis,” in:Mathematical Modeling Methods [in Russian], MGU, Moscow (1998), pp. 42–53.
N. A. Weiss,Introductory Statistics, New York (1995).
K. F. Gauss,Theoria Motus Corporum Coelstium in Sectioninus Copecis Solem Ambientium, Hamburg (1809).
A. A. Borovkov,Probability Theory [in Russian], Nauka, Moscow (1986).
Additional information
Translated from Prikladnaya Matematika i Informatika, No. 2, pp. 94–98, 1999.
Rights and permissions
About this article
Cite this article
Mechenov, A.S. Least distance method for a confluent model with homoskedasticity in the matrix columns and the right-hand side. Comput Math Model 11, 299–304 (2000). https://doi.org/10.1007/BF02361135
Issue Date:
DOI: https://doi.org/10.1007/BF02361135