Admissible estimation in an one parameter nonregular family of absolutely continuous distributions

Consider the problem of estimating under squared error loss an arbitrarily positive, strictly increasing or decreasing parametric function based on a sample of size n in an one parameter nonregular family of absolutly continuous distributions with both endpoints of the support depending on a single parameter. We first provide sufficient conditions for the admissibility of generalized Bayes estimators with respect to some specific priors and then treat several examples which illustrate the admissibility of best invariant estimators in some location or scale parameter problems.     

A note on comparison and improvement of estimators based on likelihood

The problem of estimation of a parameter of interest in the presence of a nuisance parameter, which is either location or scale, is considered. Three estimators are taken into account: usual maximum likelihood (ML) estimator, maximum integrated likelihood estimator and the bias-corrected ML estimator. General results on comparison of these estimators w.r.t. the second-order risk based on the mean-squared error are obtained. Possible improvements of basic estimators via the notion of admissibility and methodology given in Ghosh and Sinha [A necessary and sufficient condition for second order admissibility with applications to Berkson's bioassay problem. Ann Stat. 1981;9(6):1334–1338] are considered. In the recent paper by Tanaka et al. [On improved estimation of a gamma shape parameter. Statistics. 2014; doi:10.1080/02331888.2014.915842], this problem was considered for estimating the shape parameter of gamma distribution. Here, we perform more accurate comparison of estimators for this case as well as for some other cases.     

Estimation of a Gamma Scale Parameter Under Asymmetric Squared Log Error Loss

Abstract

Estimation of scale parameter under the squared log error loss function is considered with restriction to the principle of invariance and risk unbiasedness. An explicit form of minimum risk scale-equivariant estimator under this loss is obtained. The admissibility and inadmissibility of a class of linear estimators of the form (cT + d) are considered, where T follows a gamma distribution with an unknown scale parameter η and a known shape parameter ν. This includes the admissibility of the minimum risk equivariant estimator on η (MRE).     

Admissibility of linear estimators with respect to restricted parameter sets

In this paper the admissibility of a linear estimator for a linear regression parameter is characterized for such cases, where the considered parameter varies in an ellipsoid. We obtain a certain subset of the set of all linear estimators which are admissible with respect to the unrestricted parameter set. Furthermore, various linear estimators which have been proposed for improving the least squares estimator in cases of a restricted parameter set are investigated for admissibility. It turns out that only some shrunken estimators and some estimators of ridge type are admissible, whereas the KUKS-OLMAN estimator and all estimators of MARQUARDT type can be improved.     

Admissible estimators in finite population sampling employing various types of prior information

We consider some estimators of the total and variance of a finite population from Bayesian and pseudo-Bayesian perspectives. Recently, Meeden and Ghosh (1982a, 1982b) have provided quite simple but powerful tools for proving admissibility of estimators and estimator-design pairs is finite population sampling problems. We consider what these techniques yield in the way of admissibility results for the estimators discussed.     

ESTIMATION OF THE PARAMETER OF A POISSON DISTRIBUTION USING A LINEX LOSS FUNCTION

This paper considers estimation of the parameter of a Poisson distribution using Varian's (1975) asymmetric LINEX loss function L (δ) = b{exp(aδ) - aδ - 1}, where δ is the estimation error and b > 0, a 0. It is shown that for a < 0, the sample mean X¯ is admissible whereas for a > 0, X¯ is dominated by c*X¯, where c*= (n/a)log(1+a/n). Practical implications of this result are indicated. More general results, concerning the admissibility of estimators of the form cX¯+ d are also presented.     

Estimation of parameters of exponential distribution in the truncated space using asymmetric loss function

This paper is concerned with estimation of location and scale parameters of an exponential distribution when the location parameter is bounded above by a known constant. We propose estimators which are better than the standard estimators in the unrestricted case with respect to the suitable choice of LINEX loss. The admissibility of the modified Pitman estimators with respect to the LINEX loss is proved. Finally the theory developed is applied to the problem of estimating the location and scale parameters of two exponential distributions when the location parameters are ordered.     

A note on finite population prediction under asymetric loss functions

The purpose of the present note is to derive optimal population total predictors relative to the Linex (Zellner, 1986) loss function under some well known superpopulation models. The risk function and Bayes risk are derived and compared with those of usual predictors. Minimax and and admissibility properties of some of the derived predictors are also investigated.     

Admissible linear estimation in singular linear models with respect to a restricted parameter set

The admissibility results of Hoffmann (1977), proved in the context of a nonsingular covariance matrix are extended to the situation where the covariance matrix is singular. Admissible linear estimators in the Gauss-Markoff model are characterised and admissibility of the Best Linear Unbiased Estimator is investigated.     

Admissible unbiased estimation of finite population variance under a randomized response model

We consider the problem of estimation of a finite population variance related to a sensitive character under a randomized response model and prove (i) the admissibility of an estimator for a given sampling design in a class of quadratic unbiased estimators and (ii) the admissibility of a sampling strategy in a class of comparable quadratic unbiased strategies.     

An admissibility condition in the class of unbiased estimators of finite population parameters

Summary A general sufficient condition is found for estimators of a finite population parameter to be admissible in the class of its unbiased estimators. The solution extends a result given by Godambe and Joshi and appears as a unified condition which applies indistinctly to those unbiased estimators of the most usual parameters (linear and quadratic forms of the population values) for which the previous admissibility proofs were worked out separately. A further more restrictive condition proves the admissibility of estimators concerning some parameters which are non polinominal functions of the population values.     

Some remarks on an admissibility condition

Summary In this note we deal with some admissibility conditions proved by G. B. Tranquilli to be sufficient in the class of unbiased estimators of finite population parameters and with respect to (w.r.t.) a quadratic loss function. We show that the same conditions:i) are sufficient for the admissibility of an unbiased estimator with any loss function;ii) imply hyperadmissibility with reference to a particular (critical) population of the. From this fact we deduce that, for a fixed critical population, there is at most one estimator, in the class of all unbiased estimator of a finite population parameter, which satisfies Tranquilli condition. This research was partially supported by a M.U.R.S.T. grant ?Metodi inferenziali basati sul ricampionamento?.     

A note on the admissibility of relative to the linex loss function

The purpose of this note is to give a correct proof of a result in Rojo (1987). Let 2 be the mean of a random sample of size n from a normal 2 distribution with unknown mean 0 and known variance o . Following earlier work by Zellner (1986), Rojo (1987) considered the admissibility of the linear estimator c; + d relative to Variants (1975) asymmetric LINEX loss function     

Admissible Linear Estimators with Respect to Inequality Constraints

Admissibility of linear estimators is characterized in linear models E(Y)=Xβ, D(Y)=V, with an unknown multidimensional parameter (β, V) varying in the Cartesian product C × ν, where C is a subset of space and ν is a given set of non negative definite symmetric matrices. The relation between admissibility of inhomogeneous and homogeneous linear estimators is discussed, and some sufficient and necessary conditions for admissibility of an inhomogeneous linear estimator are given.     

Sufficient admissibility conditions for scale parameter estimator

Although there are a number of results available for the admissibility of the best translation equivariant estimator of the parameter, there is hardly any stated explicitly for the best scale equivariant estimator of the scale parameter. In this paper, we derive sufficient conditions for the admissibility of the scale parameter estimators and compara them. The derivations use the well known results due to Brown [1], Farrell [2], and Portnoy [3]. The loss function has been taken to be quadratic.     

ϵ-Admissible Estimators for Normal and Poisson Means

The criterion of admissibility has been considered as one of the most important criterion in decision theory and many important results have been contributed in this direction. In this article, we propose a more flexible criterion, so-called ?-admissibility (which can be considered as weak admissibility), which generates a monotone sequence of classes of estimators. The limit of this sequence, class of 0⁺-admissible estimators, is the smallest class including the class of usual admissible estimators, which also belongs to the monotone sequence. Some sufficient and necessary conditions are proposed for ?-admissibility and 0⁺-admissibility. Under some weighted square loss, it can be shown that the usual MLE is 0⁺-admissible for the multivariate normal distribution and the multivariate Poisson distribution.     

Smoothing parameter selection in nonparametric regression using an improved Akaike information criterion

Many different methods have been proposed to construct nonparametric estimates of a smooth regression function, including local polynomial, (convolution) kernel and smoothing spline estimators. Each of these estimators uses a smoothing parameter to control the amount of smoothing performed on a given data set. In this paper an improved version of a criterion based on the Akaike information criterion (AIC), termed AIC_C, is derived and examined as a way to choose the smoothing parameter. Unlike plug-in methods, AIC_C can be used to choose smoothing parameters for any linear smoother, including local quadratic and smoothing spline estimators. The use of AIC_C avoids the large variability and tendency to undersmooth (compared with the actual minimizer of average squared error) seen when other 'classical' approaches (such as generalized cross-validation (GCV) or the AIC) are used to choose the smoothing parameter. Monte Carlo simulations demonstrate that the AIC_C-based smoothing parameter is competitive with a plug-in method (assuming that one exists) when the plug-in method works well but also performs well when the plug-in approach fails or is unavailable.     

Estimating a positive normal mean

Let X ₁, X ₂, ..., X _n be a random sample from a normal population with mean μ and variance σ ². In many real life situations, specially in lifetime or reliability estimation, the parameter μ is known a priori to lie in an interval [a, ∞). This makes the usual maximum likelihood estimator (MLE) ̄ an inadmissible estimator of μ with respect to the squared error loss. This is due to the fact that it may take values outside the parameter space. Katz (1961) and Gupta and Rohatgi (1980) proposed estimators which lie completely in the given interval. In this paper we derive some new estimators for μ and present a comparative study of the risk performance of these estimators. Both the known and unknown variance cases have been explored. The new estimators are shown to have superior risk performance over the existing ones over large portions of the parameter space.     

Prediction Error Property of the Lasso Estimator and its Generalization

The lasso procedure is an estimator‐shrinkage and variable selection method. This paper shows that there always exists an interval of tuning parameter values such that the corresponding mean squared prediction error for the lasso estimator is smaller than for the ordinary least squares estimator. For an estimator satisfying some condition such as unbiasedness, the paper defines the corresponding generalized lasso estimator. Its mean squared prediction error is shown to be smaller than that of the estimator for values of the tuning parameter in some interval. This implies that all unbiased estimators are not admissible. Simulation results for five models support the theoretical results.