Abstract
The concept of a matrix cone in ℝ p n is introduced and its basic properties are studied. Based on this concept, the problem of testing conic hypotheses in a multivariate Gaussian analysis is considered. This problem generalizes the corresponding univariate analogues. The distribution of the critical test is obtained under the hypothesis.
Similar content being viewed by others
References
P. A. Kashitsyn, “Multivariate Model with a Kronecker Product Covariance Structure: S. N. Roy Method of Estimation,” Math. Methods Stat. 20 (1), 75 (2011).
P. A. Kashitsyn, “Multivariate Model with Correlated Observation Units,” Theory Probab. Appl. 56 (3), 602 (2011).
Yu. N. Tyurin, “Checking Conic Hypotheses,” in Mathematics, Mechanics, Informatics. Proc. Gonf. Dedicated to 10th Anniversary of RFBR (Fizmatlit, Moscow, 2005), pp. 289–307.
Author information
Authors and Affiliations
Corresponding author
Additional information
Original Russian Text © P.A. Kashitsyn, 2015, published in Vestnik Moskovskogo Universiteta, Matematika. Mekhanika, 2015, Vol. 70, No. 3, pp. 15–18.
About this article
Cite this article
Kashitsyn, P.A. Verification of conical hypotheses in the multivariate Gaussian analysis. Moscow Univ. Math. Bull. 70, 115–118 (2015). https://doi.org/10.3103/S0027132215030031
Received:
Published:
Issue Date:
DOI: https://doi.org/10.3103/S0027132215030031