Email this article On the asymptotic optimality of orthoregressional estimators
- Authors: Lomov A.A.1,2
-
Affiliations:
- Sobolev Institute of Mathematics
- Novosibirsk State University
- Issue: Vol 10, No 4 (2016)
- Pages: 511-519
- Section: Article
- URL: https://journals.rcsi.science/1990-4789/article/view/212513
- DOI: https://doi.org/10.1134/S1990478916040074
- ID: 212513
Cite item
Open Access
Access granted
Subscription Access
Abstract
It is shown that the orthoregressional (STLS) parameter estimators in linear algebraic systems (including autonomous difference equations with matrix coefficients) converge to the maximum likelihood estimators and thus become asymptotically best in the limit case of large variances of the random coordinates on the variety of solutions to the system observed with additive random perturbations.
About the authors
A. A. Lomov
Sobolev Institute of Mathematics; Novosibirsk State University
Author for correspondence.
Email: lomov@math.nsc.ru
Russian Federation, pr. Akad. Koptyuga 4, Novosibirsk, 630090; ul. Pirogova 2, Novosibirsk, 630090
Supplementary files
