Double Moving Extremes Ranked Set Sampling Design

The traditional simple random sampling(SRS) design method is inefficient in many cases. Statisticians proposed some new designs to increase efficiency. In this paper, as a variation of moving extremes ranked set sampling(MERSS), double MERSS(DMERSS) is proposed and its properties for estimating the population mean are considered. It turns out that, when the underlying distribution is symmetric, DMERSS gives unbiased estimators of the population mean. Also, it is found that DMERSS is more efficie...     

BAYESIAN ANALYSIS OF DATA WITH ONLY ONE FAILURE

The hearings of a certain type have their lives following a Weibull distribution. In a life test with 20 sets of bearings, only one set failed within the specified time, and none of the remainder failed even after the time of to estimate the reliabilWith a set of testing data like that in Table 1, it is required to estimate the reliability at the mission time, In this paper, we first use hierarchical Bayesian method of determine the prior distribution and the Bayesian estimates of various probabilities of failures, pi‘s, then use the method of least squares to estimate the parameters of the Weibull distribution and the reliability. Actual computation shows that the estimates so obtained are rather robust. And the results have been adopted for practical use.     

随机规划中若干凸性命题(Ⅰ)

There is a Kind of so-called chance constrained programming problems in stochastic programming. One of the main subjects of this kind of problems is whether the constrained set or is convex. In this letter, we discuss the convexity statement if all the elements of the fist column of A(ω) are independent random variables with known Weibull distributions and others are fixed. Suppose density function for Weibull distribution     

Parameter estimation of the WMTD model

The MTD （mixture transition distribution） model based on Weibull distribution （WMTD model） is proposed in this paper, which is aimed at its parameter estimation. An EM algorithm for estimation is given and shown to work well by some simulations. And bootstrap method is used to obtain confidence regions for the parameters. Finally, the results of a real example--predicting stock prices--show that the WMTD model proposed is able to capture the features of the data from thick-tailed distribution better than GMTD （mixture transition distribution） model.     

Optimal foldover plans of three-level designs with minimum wrap-around L-_2 discrepancy

Abstract The objective of this paper is to study the issue of employing the uniformity criterion measured by the wrap-around L2-discrepancy to assess the optimal foldover plans for three-level designs.For three-level fractional factorials as the original designs,the general foldover plan and combined design under a foldover plan are defined,some theoretical properties of the defined foldover plans are obtained,a tight lower bound of the wrap-around L2-discrepancy of combined designs under a general foldover plan is also obtained,which can be used as a benchmark for searching optimal foldover plans.For illustration of the usage of our theoretical results,a catalog of optimal foldover plans for uniform initial designs with s three-level factors is tabulated,where 2≤ s ≤11.     

Bayesian Approach for Recovering Piecewise Constant Viscoelasticity from MRE Data

This paper deals with an inverse problem for recovering the piecewise constant viscoelasticity of a living body from MRE(Magnetic Resonance Elastography)data.Based on a scalar partial differential equation whose solution can approximately simulate MRE data,our inverse coefficient problem is considered as a statistical inverse problem of reconstructing the posterior distribution of unknown viscoelastic modulus.For sampling this distribution,one usually can use the Metropolis-Hastings Markov chain Monte Carlo(MHMCMC)algorithm.However,without an appropriate"proposal"distribution given artificially,the MH-MCMC algorithm is hard to draw samples efficiently.To avoid this,a so-called slice sampling algorithm is introduced in this paper and applied for solving our problem.The performance of these statistical inversion algorithms is numerically tested basing on simulated data.     

Multiwavelet sampling theorem in Sobolev spaces

This paper is to establish the multiwavelet sampling theorem in Sobolev spaces. Sampling theorem plays a very important role in digital signal communication. The most classical sampling theorem is Shannon sampling theorem, which works for bandlimited signals. Recently, sampling theorems in wavelets or multiwavelets subspaces are extensively studied in the literature. In this paper, we firstly propose the concept of dual multiwavelet frames in dual Sobolev spaces (H s (R) , H-s (R)). Then we construct a special class of dual multiwavelet frames, from which the multiwavelet sampling theorem in Sobolev spaces is obtained. That is, for any f ∈ H s (R) with s 1/2, it can be exactly recovered by its samples. Especially, the sampling theorem works for continuous signals in L 2 (R), whose Sobolev exponents are greater than 1 /2.     

EXPECTED NUMBER OF ITERATIONS OF INTERIOR-POINT ALGORITHMS FOR LINEAR PROGRAMMING

We study the behavior of some polynomial interior-point algorithms for solving random linear programming (LP) problems. We show that the expected and anticipated number of iterations of theseTodd‘s probabilisticalgorithms is bounded above by O(n^1.5). The random LP problem is model with the Cauchy distribution.     

三参数Weibull分布参数和其它可靠性指标的区间估计方法

This paper discusses the interval estimations method for the parameters and other reliability characters of a three-poxameter Weibull distribution. According to the fiducial distrlbution theory of the parameter, the author presents the confidence intervals of the porameters, the reliability and the reliable life. An example mad simulation results are given. It is shown that the method presented in this paper is practicable and worth noticing.     

SAMPLING INSPECTION OF RELIABILITY IN （LOG）NORMAL CASE WITH TYPE I CENSORING

This article proposes a statistical method for working out reliability sampling plans under Type I censored sample for items whose failure times have either normal or lognormal distributions. The quality statistic is a method of moments estimator of a monotonous function of the unreliability. An approach of choosing a truncation time is recommended. The sample size and acceptability constant are approximately determined by using the Cornish-Fisher expansion for quantiles of distribution. Simulation results show that the method given in this article is feasible.     

A Weighted Likelihood Ratio of Two Related Negative Hypergeomeric Distributions

In this paper we consider some related negative hypergeometric distributions arising from the problem of sampling without replacement from an urn containing balls of different colours and in different proportions but stopping only after some specific number of balls of different colours have been obtained. With the aid of some simple recurrence relations and identities we obtain in the case of two colours the moments for the maximum negative hypergeometric distribution, the minimum negative hypergeometric distribution,the likelihood ratio negative hypergeometric distribution and consequently the likelihood proportional negative hypergeometric distributiuon. To the extent that the sampling scheme is applicable to modelling data as illustrated with a biological example and in fact many situations of estimating Bernoulli parameters for binary traits within a finite population, these are important first-step results.     

STABLE DOUBLE LR ALGORITHM AND ITS ERROR ANALYSIS

In this paper, the normative matrices and their double LR transformation with origin shifts are defined, and the essential relationship between the double LR transformation of a normative matrix and the QR transformation of the related symmetric tridiagonal matrix is proved. We obtain a stable double LR algorithm for double LR transformation of normative matrices and give the error analysis of our algorithm. The operation number of the stable double LR algorithm for normative matrices is only four sevenths of the rational QR algorithm for reed symmetric tridiagonal matrices.     

STATIONARY SOLUTIONS OF THE RELATIVISTIC VLASOV-MAXWELL SYSTEM OF PLASMA PHYSICS

The authors consider the stationary relativistic coupled system consisting of Vlasov’s equation for the distribution function of charged particles and Maxwell''s equations for the electric and magnetic fields of a plasma. With different tools of nonlinear functional analysis the existence of solutions is proved, in which, according to different geometries and symmetries, the distribution function depends on one, two or three independent integrals of the motion.     

Global Convergence of the Broyden‘’s Class of Quasi—Newton Methods with Nonmonotone Linesearch

In this paper,the Broyden class of quasi-Newton methods for unconstrained optimization is inves-tigated.Non-monotone linesearch procedure is introduced,which is combined with the Broyden‘s class.Under the convexity assumption on objective function,the global convergence of the Broyden‘s class is proved.     

THE EXPANSION FOR THE ERROR DISTRIBUTION OF SAMPLE MEDIAN AND ITS RANDOM WEIGHTING STATISTIC

In this paper,the expansion for the error distribution of the sample median is obtained.To mimic the error distribution of the sample median,random weighting statistic is introduced.It is proved in this paper that the precision of the approximation of the error distribution for sample medisn by random weighting method is of the order of o.     

Nonuniform sampling and approximation in Sobolev space from perturbation of the framelet system

The Sobolev space H~?(R~d), where ? d/2, is an important function space that has many applications in various areas of research. Attributed to the inertia of a measurement instrument, it is desirable in sampling theory to recover a function by its nonuniform sampling. In the present paper, based on dual framelet systems for the Sobolev space pair(H~s(R~d), H~(-s)(R~d)), where d/2 s ?, we investigate the problem of constructing the approximations to all the functions in H~?(R~d) by nonuniform sampling. We first establish the convergence rate of the framelet series in(H~s(R~d), H~(-s)(R~d)), and then construct the framelet approximation operator that acts on the entire space H~?(R~d). We examine the stability property for the framelet approximation operator with respect to the perturbations of shift parameters, and obtain an estimate bound for the perturbation error. Our result shows that under the condition d/2 s ?, the approximation operator is robust to shift perturbations. Motivated by Hamm(2015)'s work on nonuniform sampling and approximation in the Sobolev space, we do not require the perturbation sequence to be in ?~α(Z~d). Our results allow us to establish the approximation for every function in H~?(R~d) by nonuniform sampling. In particular, the approximation error is robust to the jittering of the samples.     

ON THE SPH-DISTRIBUTION CLASS

Following up Neuts‘ idea, the SPH-distribution class associated with bounded Q matrices for infinite Markov chains is defined. The main result in this paper is to characterize the SPH class through the derivatives of the distribution functions. Based on the characterization theorem, closure properties, the expansion, uniform approximation,and the matrix representations of the SPH class are also discussed by the derivatives of the distribution functions at origin.Key words Phase type distribution, absorbing Markov chain, operator theory, SPH- distribution, properties     

ESTIMATES OF THE DOUBLE DETERMINANTS OF QUATERNION MATRICES

An estimate of the upper bound is given for the double determinant of the sum of two arbitrary quaternion matrices, and meanwhile the lower bound on the double determinant is established especially for the sum of two quaternion matrices which form an assortive pair. As applications, some known results are obtained as corollaries and a question in the matrix determinant theory is answered completely.     

Maximum Likelihood Estimator of the Location Parameter under Moving Extremes Ranked Set Sampling Design

Cost effective sampling design is a problem of major concern in some experiments especially when the measurement of the characteristic of interest is costly or painful or time consuming.In the current paper,a modification of ranked set sampling(RSS)called moving extremes RSS(MERSS)is considered for the estimation of the location parameter for location family.A maximum likelihood estimator(MLE)of the location parameter for this family is studied and its properties are obtained.We prove that the MLE is an equivariant estimator under location transformation.In order to give more insight into the performance of MERSS with respect to(w.r.t.)simple random sampling(SRS),the asymptotic efficiency of the MLE using MERSS w.r.t.that using SRS is computed for some usual location distributions.The relative results show that the MLE using MERSS can be real competitors to the MLE using SRS.     

Huber＇s Minimax Approach in Distribution Classes with Bounded Variances and Subranges with Applications to Robust Detection of Signals

A brief survey of former and recent results on Huber‘s minimax approach in robust statistics is given. The least informative distributions minimizing Fisher information for location over several distribution classes with upper-bounded variances and subranges are written down. These least informative distributions are qualitatively different from classical Huber‘s solution and have the following common structure: (i) with relatively small variances they are short-tailed, in particular normal；(ii) with relatively large variances they are heavytailed, in particular the Laplace; (iii) they are compromise with relatively moderate variances. These results allow to raise the efficiency of minimax robust procedures retaining high stability as compared to classical Huber‘s procedure for contaminated normal populations. In application to signal detection problems, the proposed minimax detection rule has proved to be robust and close to Huber‘s for heavy-tailed distributions and more efficient than Huber‘s for short-tailed ones both in asymptotics and on finite samples。