Abstract
Two approaches to the multiplication of an uncertain vector in R n by an exactly known matrix are considered. Namely, uncertain vectors can be localized either in boxes or in ellipsoids. The accuracy of these approaches is compared by measuring the volume of the final localization domain. The main result is that, under additional assumptions, the probability of the event that the ellipsoidal localization is preferable tends to 1 as n → ∞.
Similar content being viewed by others
References
C. L. Siegel, “Über die analytische Theorie der quadratischen Formen,” Ann. Math. Ser. 2 36, 527–606 (1935).
R. Godement, Fonctions holomorphes de carré sommable dans le demi-plan de Siegel, Seminaire Henri Cartan 10(1), Exp. no. 6 (École Normale Supérieure, Paris, 1957–1958).
S. Helgason, Differential Geometry and Symmetric Spaces (Academic Press, New York-London, 1962).
A. I. Ovseevich, Yu.V. Taraban’ko, and F. L. Chernous’ko, “Comparison of Interval and Ellipsoidal Bounds for the Errors of Vector Operations,” Dokl. Akad. Nauk 400(6), 739–743 (2005) [Doklady Math. 71 (1), 127–130 (2005)].
B. Simon, Functional Integration and Quantum Physics (Academic Press, New York-San Francisco-London, 1979).
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Ovseevich, A.I. Stochastic approach to the comparison of ellipsoidal and interval estimates. Russ. J. Math. Phys. 16, 529–535 (2009). https://doi.org/10.1134/S1061920809040074
Received:
Published:
Issue Date:
DOI: https://doi.org/10.1134/S1061920809040074