右删失左截断情形下分布函数的分位数估计

文中考虑了右删失左截断数据情形下分布函数的分位数估计，讨论了该估计的渐近性质并获得了它的强弱Ｂａｈａｄｕｒ类型的表示定理。利用此Ｂａｈａｄｕｒ表示定理很容易获得该分位数估计的渐近正态性及置信区间等结果。     

左截断右删失数据分位差估计及其渐近性质

我们研究了左截断右删失数据分位差,基于左截断右删失数据乘积限构造了分位差的经验估计,同时克服经验估计的非光滑性,提出了分位数差的核光滑估计.利用经验过程理论推导出这两个估计的渐近偏差和渐近方差,并且在左截断右删失数据下研究了这两个分位差的大样本性质,获得分位差估计的相合性和渐近正态性.同时给出计算模拟以验证光滑分位差估计的表现,在均方损失的意义下模拟结果表明光滑估计比经验估计具有更好的性质.     

左截断右删失数据光滑分位估计的渐近性质

文中提出了随机左截断右删失数据下的一种光滑分位估计,推导出此光滑估计的相合性和渐近正态性,同时获得了该估计的强弱Bahadur表示定理。     

长度偏差右删失数据下分位数回归的估计方程方法

本文首先建立左截断右删失数据下的一般分位数回归方法.当截断变量服从均匀分布时,左截断右删失数据变成长度偏差右删失数据.长度偏差数据因其特殊性,提供了更多的信息.当把适用于左截断右删失数据的一般方法用到长度偏差右删失数据时,得到的估计量并不有效,这是因为它们没有利用该数据的特殊结构.为了提高效率,本文提出复合估计方程方法来解决长度偏差右删失数据下的分位数回归问题,这种方法并不需要估计删失变量的分布.所提出的估计方程可以通过一个求L_1型凸函数最小值的简单算法来求解.本文用经验过程和随机积分的技巧建立了所提出估计量的一致相合性和弱收敛性.随机模拟验证了所提出方法在有限样本时的表现,并且给出了实例分析.     

左截断右删失数据下分位密度估计的重对数律

在左截断右删失数据下，我们基于乘积限估计给出了分位密度估计， 获得了分位密度估计及其导数的重对数律。     

左截断右删失剩余寿命分位数的置信区间构造

在医学领域、可靠性分析和人寿保险市场中,剩余寿命是重要的研究范畴之一.因此,剩余寿命分位数区间的精确估计有着重要的意义.但是,在左截断和右删失同时存在的临床数据下,样本量通常很小,传统的置信区间构造方法多数不理想,而且涉及到的估计量方差的计算非常繁琐.为了避免上述困难,文章利用Jackknife-d方法构造了左截断右删失剩余寿命分位数的置信区间.同时,通过蒙特卡罗模拟和实例分析对Jackknife-d方法和传统的4种方法进行评价.模拟结果表明:小样本下,Jackknife-d方法得到的置信区间长度最短且覆盖率在大多数情况下都接近于名义水平,是剩余寿命分位数置信区间构造的一种很好的方法.     

生存分析中乘积限估计的大样本性质

生存分析中，人们关心的问题之一是利用不完全的寿命调查数据估计生物折寿命分布。在实际问题中，比较常见的不完全数据包括右删失数据，左截断数据和左截断右删失数据。利用这三种数据估计寿命分布时，常用的统计量是乘积限估计。于是，乘积限估计的大样本性质的研究一直受到关注。本文就这方面的研究近况做一比较系统的论述。     

一般偏差右删失数据下剩余寿命分位数回归

剩余寿命是刻画个体预期寿命的一个重要度量,对剩余寿命的早期研究主要集中在剩余均值上.然而当总体生存函数偏态或厚尾时剩余均值函数可能不存在,因此统计学者建议用剩余寿命分位数来刻画预期寿命.在完全数据和右删失数据下,剩余寿命分位数的建模和理论已经很完善.但是,在实际的调查研究中经常会遇到偏差抽样数据.例如,临床医学中的左截断数据,流行病学中的病例队列抽样数据,医学大型队列研究中的长度偏差抽样数据等等.忽略抽样偏差会导致参数估计有偏和不合理的推断结果.本文考虑一般偏差右删失数据下剩余寿命分位数回归的统计推断问题.首先,我们提出了一个一般偏差右删失数据下的剩余寿命分位数回归模型,并利用一般估计方程方法对模型中的参数进行了估计.针对已有文献常用的删失变量与协变量独立性假设,本文重点考虑了删失变量依赖于协变量场合.其次,由于估计量的渐近方差中涉及非参密度函数,在估计渐近方差时,本文采用Bootstrap方法.最后,数值模拟显示本文提出的方法有限样本性质表现很好.     

左截断右删失数据下光滑分布函数估计效率

研究了左截断右删失数据下光滑分布函数估计,并获得了其渐近性质.在MSE意义下,给出了光滑分布函数估计与经验估计(即乘积限估计)的相对亏量,证明了在一定的条件下,光滑分布估计要优于经验分布估计,并通过模拟说明了光滑分布函数估计比乘积限估计更加有效.     

左截断右删失下一类泛函估计的渐近性质

在左截断右删失下,本文讨论了一类广义Von－Mises泛函估计的渐近性质．在一定条件下,得到了此类泛函估计的强逼近和U-统计量表示,并由此得出它的强相合性、渐近正态性及重对数律．     

A Strong Representation of the Product-Limit Estimator for Left Truncated and Right Censored Data

In this paper we consider the TJW product-limit estimatorF_n(x) of an unknown distribution functionFwhen the data are subject to random left truncation and right censorship. An almost sure representation of PL-estimatorF_n(x) is derived with an improved error bound under some weaker assumptions. We obtain the strong approximation ofF_n(x)−F(x) by Gaussian processes and the functional law of the iterated logarithm is proved for maximal derivation of the product-limit estimator toF. A sharp rate of convergence theorem concerning the smoothed TJW product-limit estimator is obtained. Asymptotic properties of kernel estimators of density function based on TJW product-limit estimator is given.     

Sequential Confidence Bands for Quantile Densities Under Truncated and Censored Data

In this paper an asymptotic distribution is obtained for the maximal deviation between the kernel quantile density estimator and the quantile density when the data are subject to random left truncation and right censorship. Based on this result we propose a fully sequential procedure for construct ing a fixed-width confidence band for the quantile density on a finite interval and show that the procedure has the desired coverage probability asymptotically as the width of the band approaches zero.     

Quantile process for left truncated and right censored data

In this paper, we consider the product-limit quantile estimator of an unknown quantile function when the data are subject to random left truncation and right censorship. This is a parallel problem to the estimation of the unknown distribution function by the product-limit estimator under the same model. Simultaneous strong Gaussian approximations of the product-limit process and product-limit quantile process are constructed with rate . A functional law of the iterated logarithm for the maximal deviation of the estimator from the estimand is derived from the construction. Work partially supported by NSC Grant 89-2118-M-259-011.     

Statistical inference for quantiles in the frequency domain

For second-order stationary processes, the spectral distribution function is uniquely determined by the autocovariance function of the process. We define the quantiles of the spectral distribution function in frequency domain. The estimation of quantiles for second-order stationary processes is considered by minimizing the so-called check function. The quantile estimator is shown to be asymptotically normal. We also consider a hypothesis testing for quantiles in frequency domain and propose a test statistic associated with our quantile estimator, which asymptotically converges to standard normal under the null hypothesis. The finite sample performance of the quantile estimator is shown in our numerical studies.     

Strong approximations of the quantile process of the product-limit estimator

The quantile process of the product-limit estimator (PL-quantile process) in the random censorship model from the right is studied via strong approximation methods. Some almost sure fluctuation properties of the said process are studied. Sections 3 and 4 contain strong approximations of the PL-quantile process by a generalized Kiefer process. The PL and PL-quantile processes by the same appropriate Kiefer process are approximated and it is demonstrated that this simultaneous approximation cannot be improved in general. Section 5 contains functional LIL for the PL-quantile process and also three methods of constructing confidence bands for theoretical quantiles in the random censorship model from the right.     

Improved confidence intervals for quantiles

We derive the Edgeworth expansion for the studentized version of the kernel quantile estimator. Inverting the expansion allows us to get very accurate confidence intervals for the pth quantile under general conditions. The results are applicable in practice to improve inference for quantiles when sample sizes are moderate.     

Fixed Design Nonparametric Regression with Truncated and Censored Data

In this paper we consider a fixed design model in which the observations axe subject to left truncation and right censoring. A generalized product-limit estimator for the conditional distribution at a given covaxiate value is proposed, and an almost sure asymptotic representation of this estimator is established. We also obtain the rate of uniform consistency, weak convergence and a modulus of continuity for this estimator.Applications include trimmed mean and quantile function estimators.     

Bahadur-Kiefer theorems for the product-limit process

In the random censorship from the right model, strong and weak limit theorems for Bahadur-Kiefer type processes based on the product-limit estimator are established. The main theorm is sharp and may be considered as a final result as far as this type of research is concerned. As a consequence of this theorem a sharp uniform Bahadur representation for product-limit quantiles is obtained.     

Quantile and tolerance-interval estimation in simulation

This paper discusses two sequential procedures to construct proportional half-width confidence intervals for a simulation estimator of the steady-state quantile and tolerance intervals for a stationary stochastic process having the (reasonable) property that the autocorrelation of the underlying process approaches zero with increasing lag. At each quantile to be estimated, the marginal cumulative distribution function must be absolutely continuous in some neighborhood of that quantile with a positive, continuous probability density function. These algorithms sequentially increase the simulation run length so that the quantile and tolerance-interval estimates satisfy pre-specified precision requirements. An experimental performance evaluation demonstrates the validity of these procedures.