zbMATH — the first resource for mathematics

Bayes, admissible, and minimax linear estimators in linear models with restricted parameter space. (English) Zbl 0686.62019
Summary: We consider a linear model E Y\(=X\alpha\), where the p-dimensional parameter vector \(\alpha\) is restricted by \(A\alpha\in \Omega\). The matrix A is (q\(\times p)\) of rank q, and \(\Omega\) is a compact subset of \({\mathbb{R}}^ q\), which contains zero and is symmetric to zero. The dispersion matrix of Y is known and positive definite. For estimation of a given s-dimensional linear parametric function, we confine to the class of all linear estimators with bounded mean squared error functions. The Bayes estimators within this class are found, and they are shown to coincide with the admissible estimators within this class. The minimax estimator is characterized as the linear Bayes estimator against a least favourable prior distribution, and a result helpful to find a least favourable prior is proved.
Applications are given firstly to the case of full parameter restrictions \((q=p)\), and some new results on linear admissible and linear minimax estimation are obtained. Secondly we consider the case of k unrestricted and p-k restricted parameters and estimation of unrestricted parameters. In particular the question is answered, when the BLUE from the simplified model, which ignores the restricted parameters, is the linear minimax estimator under the present model.

62F15 Bayesian inference
62J05 Linear regression; mixed models
62H12 Estimation in multivariate analysis
62F10 Point estimation
62C15 Admissibility in statistical decision theory
62C20 Minimax procedures in statistical decision theory
Full Text: DOI
[1] DOI: 10.2307/2282542 · Zbl 0116.36804 · doi:10.2307/2282542
[2] DOI: 10.1214/aos/1176346407 · Zbl 0554.62063 · doi:10.1214/aos/1176346407
[3] Drygas H., Linear Statistical Inference, Proceedings Poznan 1984 pp 48– (1985)
[4] Gaffke N., Robustness of Statustical Methods and Nonparametric Statistics, VEB Deutscher Verlag der Wissenschaften pp 37– (1984) · doi:10.1007/978-94-009-6528-7_10
[5] Gaffke N., Modern Applied Mathematics, Optimization and Operations Research pp 595– (1982)
[6] Gladitz J., Math. Operationsforsch. u. Statist., Ser. Statist 13 pp 371– (1982)
[7] Hoffmann K., Math. Operationsforsch. u. Statist., Ser. Statist 8 pp 425– (1977)
[8] Hoffmann K., Math. Operationsforsch. u. Statist., Ser. Statist 10 pp 19– (1979)
[9] DOI: 10.1016/0024-3795(83)90154-4 · Zbl 0504.15013 · doi:10.1016/0024-3795(83)90154-4
[10] Kuks J., Inv. Akad. Nauk Eston SSR 21 pp 73– (1972)
[11] Lauter H., Math. Operationsforsch. u. Statist 6 pp 689– (1975) · Zbl 0331.62048 · doi:10.1080/02331887508801248
[12] DOI: 10.1214/aos/1176345707 · Zbl 0485.62070 · doi:10.1214/aos/1176345707
[13] DOI: 10.1016/0378-3758(82)90083-0 · Zbl 0481.62056 · doi:10.1016/0378-3758(82)90083-0
[14] DOI: 10.1214/aos/1176345792 · Zbl 0489.62065 · doi:10.1214/aos/1176345792
[15] DOI: 10.1016/0378-3758(86)90141-2 · Zbl 0602.62054 · doi:10.1016/0378-3758(86)90141-2
[16] DOI: 10.1016/0378-3758(80)90020-8 · Zbl 0472.62079 · doi:10.1016/0378-3758(80)90020-8
[17] Rao C.R., Generalized Inverse of Matrices and its Applications (1971) · Zbl 0236.15004
[18] DOI: 10.1214/aos/1176344315 · Zbl 0384.62058 · doi:10.1214/aos/1176344315
[19] Vogel W., Lineares Optimieren.Akad (1967)
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.