Showing papers in "Statistics & Probability Letters in 1994"

TL;DR: In this article, the authors show that expectiles and certain M-quantiles of a distribution F are precisely the ordinary quantiles of distributions G related by an explicit formula to F.

TL;DR: In this article, a general matrix formula for computing the bias of the exact unconditional maximum likelihood estimate in ARMA models, with known and unknown mean, up to order 1/n, where n is the length of the series.

TL;DR: In this paper, the authors proposed two estimators of the entropy of a distribution function by modifying the estimator proposed by Vasicek (1976) and showed that the proposed estimators have less bias and have less mean squared error than the original estimator.

TL;DR: In this article, it was shown that the usual method of establishing Cramer's estimate also works for Levy processes, and that the Cramer estimator also works well for the Levy process.

TL;DR: In this article, a global measure, using ϕ-divergence of the posterior distributions and its curvature, is introduced to measure Bayesian robustness of classes of contaminated priors.

TL;DR: In this paper, the authors find it convenient to compute covariances, even for order statistics, by using Stein's (1972) identities, and generalize Siegel's formula to other order statistics as well as other distributions.

TL;DR: In this article, a simple immigration-birth-death process is considered, which is influenced by total catastrophes that are introduced with constant rate, and the stationary probabilities are derived using a renewal argument.

TL;DR: In this article, the estimation of the probability density function of stationary mixing processes using wavelet orthonormal bases was studied and precise asymptotic expressions for the mean integrated square estimation error were derived.

TL;DR: In this paper, a multiplicative bias reducing estimator (MBRE) was proposed for nonparametric regression. But their estimator has a pointwise convergence rate n4/9 when positive kernels are used.

TL;DR: In this paper, the authors considered the problem of finding the number of values of n ϵ {1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14, 15, 16, 17, 18, 20] for which nj = Mn.

TL;DR: In this article, the authors define a family of continuous time fractional ARMA processes and prove several probabilistic results concerning the memory of these processes and the regularity properties of their sample functions.

TL;DR: In this article, necessary and sufficient conditions for a distribution function to belong to the max domain of attraction of a max stable law under power normalization in terms of the integrals of the upper tail of the distribution function are derived.

TL;DR: In this paper, the bias curves of k-step M-estimators were derived in the location framework and in the scale case, where the bias increases only slightly with k.

TL;DR: The influence functions for Rousseeuw's (1987) least trimmed squares estimator and Tableman's (1994) least-truned absolute deviations estimator are derived in the univariate case as discussed by the authors.

TL;DR: In this article, it was shown that if the region is convex, then a dual problem always exists which is frequently more tractable than the original problem and conversely, the form of the dual problem suggests an iterative algorithm for solving a MLE problem when the constraint region can be written as a finite intersection of 'nice' constraint regions.

TL;DR: This paper gives asymptotically best data based choices of the bandwidth of the kernel density estimator in the spirit of the usual Fisher Information, with the use of only nonnegative kernel estimators at all stages of the selection process.

TL;DR: The Stein variance estimator is used to improve the James-Stein estimator of a multivariate normal mean vector as discussed by the authors, and it is also demonstrated that this domination does not hold for the positive part versions of these estimators.

TL;DR: In this article, it was shown that strong unimodality is not always a sufficient condition for the optimal cutpoint to occur at the mean, and that splitting a symmetric, unimodal distribution into two groups so as to minimize the sum of the within-group variances does not always occur.

TL;DR: In this article, the least trimmed absolute deviations (LTAD) estimator is formally treated and compared with the LMS and LTS estimator. And the asymptotic properties of the LTAD estimator are derived.

TL;DR: In this article, Rank statistics and sign statistics are used to study the properties of the simple random walk for the change-point problem, when the alternative is of the epidemic or square-wave type.

TL;DR: In this article, the autocorrelation parameter σ is studied in nonparametric and semiparametric regression models with autoregressive errors and it is shown that under mild assumptions one can construct an estimator that is asymptotically equivalent to the least squares estimator based on the autoregression error process.

TL;DR: In this article, a general class of goodness-of-fit tests called disparity tests containing the family of power weighted divergence statistics as a subclass is considered and the asymptotic distribution of disparity tests is shown to be chi-square.

TL;DR: In this paper, it was shown that M-estimators of scale have a rather low efficiency in the location model and that this is also the case for M-stimulators of scale at normal models.

TL;DR: In this article, a review of positive-breakdown regression is presented, where several unusual aspects are shown to be intimately connected with the exact fit property, and it is argued that the latter is not a drawback but an interesting property which helps to explain why positive breakdown estimators often succeed at revealing a hidden structure in the data.

TL;DR: In this paper, conditions on the counting process such as the mean residual life ordering, the increasing convex ordering and the expectation ordering between M1 and M2 are preserved in the random lifetimes of the two devices.

TL;DR: In this paper, a method for very fast evaluation of the inverse of the normal distribution is presented. But it is not suitable for generating a normal random variable by direct inversion of its distribution function.

TL;DR: In this article, the authors illustrate the principle via the derivation of the restricted (or residual) (REML) estimating equations, which can be obtained under more general conditions using quasi-likelihood methods.

TL;DR: Simes' (1986) procedure for testing an intersection null hypothesis in terms of the individual p -values corresponding to the component hypotheses is extended to allow the allotment of different weights to the different component hypotheses as discussed by the authors.

TL;DR: The limit distribution of the difference between the empirical distribution and its least concave majorant at a fixed point is derived in this article, where the limit distribution is defined as the sum of the two distributions at the same time.

TL;DR: In this article, a new test statistic is proposed for testing goodness of fit of an m th order polynomial regression model, which is ∫ 1 0 [μ ( m ) λ ( t )] 2 d t.