Convergence of a fixed point algorithm

This note gives a condition which ensures convergence to an optimum of an algorithm suggested by Silvey, Titterington and Torsney (19 78) in the optimal design context.     

Asymptotic results for model robust regression

Since the mid 1980's many statisticians have studied methods for combining parametric and nonparametric models to improve the quality of fits in a regression problem. Notably Einsporn (1987) proposed the Model Robust Regression 1 estimate (MRRl) in which the parametric function, f, and the nonparametric functiong were combined in a straightforward fashion via the use of a mixing parameter, λ This technique was studied extensively atsmall samples and was shown to be quite effective at modeling various unusual functions. In this paper we have asymptotic results for the MRRl estimate in the case where λ is theoretically optimal, is asymptotically optimal and data driven, and is chosen with the PRESS statistic (Allen, 1971) We demonstrate that the MRRl estimate with λchosen by the PRESS statistic is slightly inferior asymptotically to the other two estimates, but, nevertheless possesses positive asymptotic qualities.     

Moving truncations barrier-function methos for estimation in three-parameter lognormal models

A method of centres algorithm for maximum likelihood estimation in the three-parameter lognormal model is presented and discussed, The algorithm is a member of the class of moving truncations algorithms for solving nonlinear programming problems and is able to move the numerical search out of the region of the infinite maximum of the conditional likelihood function, thereby permitting convergence to an interior relative maximum of this function. The algorithm also includes an optimality test to locate the primary relative maximum of the likelihood function.     

On estimating the difference of location parameters of two negative exponential distributions

Fixed-width confidence intervals for the difference of location parameters of two negative-exponential distributions have been constructed through two-stage and purely sequential schemes. The two cases when the scale parameters are equal but unknown, and unequal but unknown, have been dealt with separately. Our two-stage procedures guarantee the exact confidence coefficient to be at least the nominal prescribed level. Various second-order expansions are also considered when sequential procedures are proposed. It is noted that no new tables are needed to implement these procedures in practice.     

A method for approximate fully bayesian analysis of parameters

Stochastic modeling of the geology in petroleum reservoirs has become an important tool in order to investigate flow properties in the reservoir. The stochastic models used contain parameters which must be estimated based on observations and geological knowledge. The amount of data available is however quite limited due to high drilling costs etc., and the lack of data prevents the use of many of the standard data driven approaches to the parameter estimation problem. Modern simulation based methods using Markov chain Monte Carlo simulation, can however be used to do fully Bayesian analysis with respect to parameters in the reservoir model, with the drawback of relatively high computational costs. In this paper, we propose a simple, relatively fast approximate method for fully Bayesian analysis of the parameters. We illustrate the method on both simulated and real data using a two-dimensional marked point model for reservoir characterization.     

Generalised partial linear single-index mixed models for repeated measures data

In this paper, we propose generalised partial linear single-index mixed models for analysing repeated measures data. A penalised quasi-likelihood approach using P-spline is used to estimate the nonparametric function, linear parameters, and single-index coefficients. Asymptotic properties of the estimators are developed when the dimension of spline basis grows with increasing sample size. Simulation examples and two applications: the study of health effects of air pollution in North Carolina, and treatment effect of naltrexone on health costs for alcohol-dependent individuals, illustrate the effectiveness of our approach.     

Robustness of automated data choices of smoothing parameter in image regularization

Statistical image restoration techniques are oriented mainly toward modelling the image degradation process in order to recover the original image. This usually involves formulating a criterion function that will yield some optimal estimate of the desired image. Often these techniques assume that the point spread function is known when the image is restored and indeed when we estimate the smoothing parameter. However in practice this assumption may not hold. This paper investigates empirically the effect of mis-specifying the point spread function on some data-based estimates of the regularization parameter and hence on the image reconstructions. Comparisons of image reconstruction quality are based on the mean absolute difference in pixel intensities between the true and reconstructed images.     

Asymptotic confidence interval of power spectrum of a continuous time process through progressively faster sampling

If the power spectral density of a continuous time stationary stochastic process is not limited to a finite bandwidth, data sampled from that process at any uniform sampling rate leads to biased and inconsistent spectrum estimators, which are unsuitable for constructing confidence intervals. In this paper, we use the smoothed periodogram estimator to construct asymptotic confidence intervals shrinking to the true spectra, by allowing the sampling rate to go to infinity suitably fast as the sample size goes to infinity. The proposed method requires minimal computation, as it does not involve bootstrap or other resampling. The method is illustrated through a Monte-Carlo simulation study, and its performance is compared with that of the corresponding method based on uniform sampling at a fixed rate.     

Automatic and asymptotically optimal data sharpening for nonparametric regression

In this article we consider data-sharpening methods for nonparametric regression. In particular modifications are made to existing methods in the following two directions. First, we introduce a new tuning parameter to control the extent to which the data are to be sharpened, so that the amount of sharpening is adaptive and can be tuned to best suit the data at hand. We call this new parameter the sharpening parameter. Second, we develop automatic methods for jointly choosing the value of this sharpening parameter as well as the values of other required smoothing parameters. These automatic parameter selection methods are shown to be asymptotically optimal in a well defined sense. Numerical experiments were also conducted to evaluate their finite-sample performances. To the best of our knowledge, there is no bandwidth selection method developed in the literature for sharpened nonparametric regression.     

ASYMPTOTIC DISTRIBUTIONS OF SEMIPARAMETRIC MAXIMUM LIKELIHOOD ESTIMATORS WITH ESTIMATING EQUATIONS FOR GROUP-CENSORED DATA

Semiparametric maximum likelihood estimation with estimating equations (SMLE) is more flexible than traditional methods; it has fewer restrictions on distributions and regression models. The required information about distribution and regression structures is incorporated in estimating equations of the SMLE to improve the estimation quality of non‐parametric methods. The likelihood of SMLE for censored data involves complicated implicit functions without closed‐form expressions, and the first derivatives of the log‐profile‐likelihood cannot be expressed as summations of independent and identically distributed random variables; it is challenging to derive asymptotic properties of the SMLE for censored data. For group‐censored data, the paper shows that all the implicit functions are well defined and obtains the asymptotic distributions of the SMLE for model parameters and lifetime distributions. With several examples the paper compares the SMLE, the regular non‐parametric likelihood estimation method and the parametric MLEs in terms of their asymptotic efficiencies, and illustrates application of SMLE. Various asymptotic distributions of the likelihood ratio statistics are derived for testing the adequacy of estimating equations and a partial set of parameters equal to some known values.     

Least-squares estimation of transition probabilities from aggregate data

Consider a population of n individuals that move independently among a finite set {1, 2,……, k} of states in a sequence of trials. t = 0. 1, 2,…, m. each according to a Markov chain with transition probability matrix P . This paper deals with the problem of estimating P on the basis of aggregate data which record only the numbers of individuals that occupy each of the k states at times t = 0. 1,2,……,m. Estimation is accomplished using conditional least squares, and asymptotic results are verified for the case n → ∞. A weighted least-squares estimator is introduced and compared with previous estimators. Some comments are made on estimability questions that arise when only aggregate data are available.     

Estimation for change point of discretely observed ergodic diffusion processes

We treat the change point problem in ergodic diffusion processes from discrete observations. Tonaki et al. (2021a) proposed adaptive tests for detecting changes in the diffusion and drift parameters in ergodic diffusion process models. When any change in the diffusion or drift parameter is detected by this or any other method, the next question to consider is where the change point is located. Therefore, we propose the method to estimate the change point of the parameter for two cases: the case where there is a change in the diffusion parameter, and the case where there is no change in the diffusion parameter but a change in the drift parameter. Furthermore, we present rates of convergence and distributional results of the change point estimators. Some examples and simulation results are also given.     

Uniform asymptotics for ruin probability of a two-dimensional dependent renewal risk model

ABSTRACT

In this paper we consider the tail behavior of a two-dimensional dependent renewal risk model with two dependent classes of insurance business, in which the claim sizes are governed by a common renewal counting process, and their inter-arrival times are dependent, identically distributed. For the case that the claim size distribution belongs to the intersection of long-tailed distribution class and dominant variation class, we obtain an asymptotic formula, which holds uniformly for all time in an infinite interval. Moreover, we point out that the formula still holds uniformly for all time in an infinite interval for widely dependent random variables (r.v.s) under some conditions.     

What is the effective sample size of a spatial point process?

Point process models are a natural approach for modelling data that arise as point events. In the case of Poisson counts, these may be fitted easily as a weighted Poisson regression. Point processes lack the notion of sample size. This is problematic for model selection, because various classical criteria such as the Bayesian information criterion (BIC) are a function of the sample size, n, and are derived in an asymptotic framework where n tends to infinity. In this paper, we develop an asymptotic result for Poisson point process models in which the observed number of point events, m, plays the role that sample size does in the classical regression context. Following from this result, we derive a version of BIC for point process models, and when fitted via penalised likelihood, conditions for the LASSO penalty that ensure consistency in estimation and the oracle property. We discuss challenges extending these results to the wider class of Gibbs models, of which the Poisson point process model is a special case.     

Nonuniqueness of least absolute values regression

The question of nonuniqueness of the least absolute values (L₁) regression is discussed, and examples are given of situations where the L₁-regression is unique and where it has 2, 3 and 4 limiting positions. A method is proposed for finding a compromise regression line when the L₁ regression is not unique. Suggestions are made for further research.     

有限总体基尼系数的统计推断

进行基尼系数的统计推断时,已有研究一般都设定总体是无限的。研究在有限总体随机抽样的现实背景下,如何利用样本数据对总体基尼系数进行估计及其评价。从总体基尼系数的内涵和定义出发,介绍了非参数估计和参数估计的方法,构造了相应的估计量,基于蒙特卡罗的模拟结果,论证和揭示了不同情况下估计量的性质,并论述了方法的适用性以及在实际应用中需要注意的问题。     

Maximum entropy estimation for survey sampling

Calibration methods have been widely studied in survey sampling over the last decades. Viewing calibration as an inverse problem, we extend the calibration technique by using a maximum entropy method. Finding the optimal weights is achieved by considering random weights and looking for a discrete distribution which maximizes an entropy under the calibration constraint. This method points a new frame for the computation of such estimates and the investigation of its statistical properties.     

On accelerating sequential procedures for estimating exponential populations

In this study, we accelerate the purely sequential procedure due to Anscombe(1953), Chow and Robbins(1965) to reduce the number of sampling operations required to carry out the estimation process. The method is proposed while estimating the location parameter(s) of the exponential distribution(s). We also develop theory for the asymptotic characteristic of the associated stopping variables. Our findings are applicable to both point as well as confidence interval estimation problems. Other interesting results are also given.     

E-Bayesian estimation and its E-posterior risk of the exponential distribution parameter based on complete and type I censored samples

Abstract

This article studies E-Bayesian estimation and its E-posterior risk, for failure rate derived from exponential distribution, in the case of the two hyper parameters. In order to measure the estimated risk, the definition of E-posterior risk (expected posterior risk) is proposed based on the definition of E-Bayesian estimation. Moreover, under the different prior distributions of hyper parameters, the formulas of E-Bayesian estimation and formulas of E-posterior risk are given respectively, these estimations are derived based on a conjugate prior distribution for the unknown parameter under the squared error loss function. Monte Carlo simulations are performed to compare the performances of the proposed methods of estimation and a real data set have been analyzed for illustrative purposes, results are compared on the basis of E-posterior risk.