Stavros Kourouklis
University of Patras
Network
Latest external collaboration on country level. Dive into details by clicking on the dots.
Publication
Featured researches published by Stavros Kourouklis.
Journal of Statistical Planning and Inference | 1998
George Iliopoulos; Stavros Kourouklis
Abstract A confidence interval for the generalized variance of a matrix normal distribution with unknown mean is constructed which improves on the usual minimum size (i.e., minimum length or minimum ratio of endpoints) interval based on the sample generalized variance alone in terms of both coverage probability and size. The method is similar to the univariate case treated by Goutis and Casella (Ann. Statist. 19 (1991) 2015–2031).
Communications in Statistics-theory and Methods | 2002
Constantinos Petropoulos; Stavros Kourouklis
ABSTRACT Under quadratic loss, a class of improved estimators for the scale parameter of an exponential distribution with unknown location is constructed. The method is analogous to Maruyamas (Minimax Estimators of a Normal Variance. Metrika 1998, 48, 209–214) construction for the variance of a normal distribution.
Annals of the Institute of Statistical Mathematics | 2001
Constantinos Petropoulos; Stavros Kourouklis
Estimation of the quantile μ + κσ of an exponential distribution with parameters (μ, σ) is considered under an arbitrary strictly convex loss function. For κ obeying a certain condition, the inadmissibility of the best affine equivariant procedure is established by exhibiting a better estimator. The LINEX loss is studied in detail. For quadratic loss, sufficient conditions are given for a scale equivariant estimator to dominate the best affine equivariant one and, when κ exceeds a lower bound specified below, a new minimax estimator is identified.
Statistics and Risk Modeling | 2000
George Iliopoulos; Stavros Kourouklis
Confidence intervals for the ratio of scale parameters are constructed in general families of distributions with nuisance (location) parameters. Each of these intervals has coverage probability at least as large as that of the standard minimum size (i.e., minimum ratio of endpoints) interval and, in addition, smaller size. Then analogous improved confidence intervals for the scale parameters subject to order restriction are derived. The method of construction is similar to that in Goutis and Casella [5], [6]. Examples are given and include the normal and exponential distributions as well as the inverse Gaussian distribution which is not a purely location–scale model. Applications to interval estimation of the error variance in variance components models are also discussed.
Statistics & Probability Letters | 1997
Stavros Kourouklis
A monotone likelihood ratio property is shown to hold for the inverse Gaussian distribution. Applications of this property in decision theoretic point and interval estimation of the lambda parameter are indicated.
Annals of the Institute of Statistical Mathematics | 1996
Stavros Kourouklis
Stein-type and Brown-type estimators are constructed for general families of distributions which improve in the sense of Pitman closeness on the closest (in a class) estimator of a parameter. The results concern mainly scale parameters but a brief discussion on improved estimation of location parameters is also included. The loss is a general continuous and strictly bowl shaped function, and the improved estimators presented do not depend on it, i.e., uniform domination is established with respect to the loss. The normal and inverse Gaussian distributions are used as illustrative examples. This work unifies and extends previous relevant results available in the literature.
Journal of Statistical Planning and Inference | 1995
Stavros Kourouklis
We consider the problem of estimating a power of the scale parameter of an exponential distribution with unknown location parameter under Pitmans measure of closeness. The loss function is assumed to be continuous and strictly bowl-shaped but is otherwise arbitrary. The optimal estimator in the class of location-scale equivariant estimators is found, and then estimators dominating it are derived by adapting Steins (1964) and Browns (1968) techniques to the present context. None of these estimators depends on the specific functional form of the loss, i.e., uniform domination with respect to the loss is established. Besides, the estimators derived from the former technique admit an interpretation as likelihood ratio testimators.
Statistics & Probability Letters | 1992
W.C.M. Kallenberg; Stavros Kourouklis
At several places in the literature there are indications that many tests are optimal in the sense of Hodges-Lehmann efficiency. It is argued here that shrinkage of the acceptance regions of the tests to the null set in a coarse way is already enough to ensure optimality. This type of argument can be used to show optimality of e.g. Kolmogorov-Smirnov tests, Cramer-von Mises tests, and likelihood ratio tests and many other tests in exponential families.
The American Statistician | 2012
Stavros Kourouklis
In 2005, Yatracos constructed the estimator S 2 2 = c 2 S 2, c 2 = (n + 2)(n − 1)[n(n + 1)]− 1, of the variance, which has smaller mean squared error (MSE) than the unbiased estimator S 2. In this work, the estimator S 2 1 = c 1 S 2, c 1 = n(n − 1)[n(n − 1) + 2]− 1, is constructed and is shown to have the following properties: (a) it has smaller MSE than S 2 2, and (b) it cannot be improved in terms of MSE by an estimator of the form cS 2, c > 0. The method of construction is based on Stein’s classical idea brought forward in 1964, is very simple, and may be taught even in an undergraduate class. Also, all the estimators of the form cS 2, c > 0, with smaller MSE than S 2 as well as all those that have the property (b) are found. In contrast to S 2, the method of moments estimator is among the latter estimators.
Journal of Multivariate Analysis | 1999
George Iliopoulos; Stavros Kourouklis