用Gibbs抽样算法计算定数截尾时Weibull分布的贝叶斯估计

本文假设产品寿命服从两参数 Weibull分布 ,在定数截尾寿命试验的情况下 ,利用近年来发展起来的 Gibbs抽样算法 ,设计了计算参数贝叶斯估计的 Gibbs抽样方案 ,计算了一个实例并与经典的 BL UE和 BL IE估计结果进行比较。结果表明 ,较之传统的数值积分法 ,Gibbs抽样算法使用起来方便直接 ,更适合于计算可靠性指标的贝叶斯估计     

关于元件失效率等级鉴定试验的一种贝叶斯方法

为鉴定元件失效率的等级,按目前国家标准GB1772-79规定的抽样方案,需要做大量的试验,构成人力、物力的浪费,做较高等级的鉴定甚至是不可能的.本文采用贝叶斯方法,利用大量的历史数据和工程技术人员的丰富经验确定先验分布,制定出元件失效率等级鉴定试验的一种新方案.按照这一方案,在确保产品质量的前提下,将大大节省试验时间,从而得到十分可观的经济效益。     

基于贝叶斯平滑转移模型的汇率非线性协整关系研究

针对平滑转移模型参数估计不确定性导致的协整检验方法相对复杂问题,提出基于平滑转移模型的贝叶斯非线性协整分析。通过模型的统计结构分析,选择参数先验分布,结合参数的后验条件分布特征设计Metropolis-Hasting-Gibbs混合抽样方案,据此估计平滑转移模型的参数,并对回归残差进行贝叶斯单位根检验,解决参数估计过程中遇到的参数估计不确定性及协整检验复杂的问题;利用人民币对美元汇率与中美两国的利率数据进行实证分析。研究结果表明:MH-Gibbs抽样方案能够有效估计平滑转移模型的参数,中美汇率波动和利差之间存在平滑转移协整关系。     

链型抽样方案综述(Ⅱ)

徐凤琴．链型抽样方案综述（Ⅱ）．本文讨论了更一般形式的链型抽样方案,称为两阶段链型抽样方案。这种方案在不同阶段使用不同的抽样比率,因而须借于平均样本量ＡＳＮ与标准型方案进行比较     

复杂机电产品多元FT VSI贝叶斯控制图性能研究

针对复杂机电产品质量控制与预警困难的问题,构建一种将多元贝叶斯统计方法与经济性能分析相结合的多元贝叶斯控制图性能分析模型.在求解模型中,多元贝叶斯控制图采用固定时间变抽样区间(Fixed Time Variable Sample Interval, FT VSI)策略,若非随机故障小概率发生,则选择宽松抽样方案;若非随机故障大概率发生,则选择严抽样方案.为量化多元贝叶斯控制图经济性能与统计性能的相关度,利用蒙卡罗模拟分析的质量控制模型进行仿真,并在不同经济性参数下,得到采样单位平均数(Average Number of Observations to Signals or End of the production run, ANOSE)对于控制图统计性能的影响程度,进而引出多元贝叶斯控制图的质量控制成本与其误报率的影响程度,并以某型号汽车自动变速器多元质量控制过程为例对多元FT VSI贝叶斯控制图性能评价与优化成果进行验证,结果证明该方法具有较好的应用性.     

一般寿命分布和定时截尾的Bayes变量抽样方案

林（1994）研究了指数分布和定时截尾的变量抽样方案．本文将讨论一般寿命分布和定时截尾的一次抽样方案．在多项式损失函数的假设下,我们讨论了Weibull分布、双参数指数分布和-分布三种情形,并着重讨论Weibull分布的情形．本文还提出了一个可用于近似地确定最优抽样方案的有报算法,并且进行了灵敏度分析,还同林较早的模型（1990,1994）做了比较．     

指数分布下Bayes鉴定试验方案

在确保使用方利益的前提下，从最大后验风险的角度运用贝叶斯方法制定了可靠性鉴定试验方案。按照此鉴定试验方案，缩短了试验时间，从而降低了鉴定试验所需的费用。     

从后验概率的应用看Bayes概率的意义

某厂有一条自动化生产线，根据以前的经验知它正常运转的概率是９５％。正常运转时生产９０％的合格品，不正常运转时可生产４０％的合格品。某日进行抽样检查，先抽取了一件产品，检验后发现它是合格品；又抽取了一件产品，检验后发现它是废品。根据两次抽样结果推测一下，生产线属于正常运转的概率有多大。这个问题中生产线正常运转的先验概率是９５％，现在产生了新的信息，即两次抽样中一次是合格品，一次是废品，在获取的这种新信息下，再推算生产线正常运转的概率即是后验概率，后验概率可用Ｂａｙｅｓ公式来计算。用Ａ表示生产线正常…     

关于MIL—HDBK—781中的保证试验

美国军用标准 MIL— H DBK— 781中的保证试验以其所需的试验时间短而引起可靠性工程界的广泛关注。本文指出这种试验方案存在不合理之处 ,并提出了避免不合理因素的修正方案 ,还从贝叶斯统计角度论证了这种修正方案是有道理的 ,值得实际工作中采用。     

关于电度表抽样检验方案

本文研究了美国有关电度表的抽样检测方案,获得了其制订抽样方案的原理,并基于此给出了几种不同批量下的抽样方案     

基于混合Ⅱ型删失数据的Weibull模型精确推断和可接受抽样计划(英文)

本文考虑基于混合Ⅱ型删失数据的Weibull模型精确推断和可接受抽样计划.得到威布尔分布未知参数最大似然估计的精确分布以及基于精确分布的置信区间.由于精确分布函数较为复杂,给出未知参数的另外几种置信区间,基于近似方法的置信区间.为了评价本文的方法,给出一些数值模拟的结果.且讨论了可靠性中的可接受抽样计划问题.利用参数最大似然估计的精确分布,给出一个可接受抽样计划的执行程序和数值模拟结果.     

随机截尾试验变量抽样接收方案的贝叶斯决策

本运用Bayes决策理论研究指数分布和随机截尾试验的抽样接收方案的一般模型,我们证明了最优Bayes法则具有单调性,并对二个特殊的决策损失函数给出了最优Bayes法则和Bayes风险的具体表达式。     

Optimal truncated repetitive lot inspection with defect rates

Single and repetitive sampling schemes are conventional methods for evaluating the quality of lots or batches of products. Truncation repetitive sampling plans are introduced in this paper in order to significantly reduce the size of samples drawn from the lot. Under this scheme type, if a decision about the acceptance or rejection of the lots cannot be made in the original inspection, they can be reinspected, at most, a prefixed number of times. The Poisson distribution is assumed for the number of defects found in the samples drawn from the lot. Best truncated repetitive sampling plans are determined by solving integer nonlinear programming problems. A simplified methodology is suggested to obtain the plans with optimal inspection effort and controlled risks by using an iterative procedure. According to the results obtained, optimal truncated plans are shown to be better than the optimal single and repetitive schemes in reducing the average sample number of the inspection. An application to the manufacturing of glass is presented for illustrative purposes.     

Some numerical quadrature schemes of a non-conforming quadrilateral finite element

Numerical quadrature schemes of a non-conforming finite element method for general second order elliptic problems in two dimensional (2-D) and three dimensional (3-D) space are discussed in this paper. We present and analyze some optimal numerical quadrature schemes. One of the schemes contains only three sampling points, which greatly improves the efficiency of numerical computations. The optimal error estimates are derived by using some traditional approaches and techniques. Lastly, some numerical results are provided to verify our theoretical analysis.     

Estimation of Binary Markov Random Fields Using Markov chain Monte Carlo

This article compares three binary Markov random fields (MRFs) which are popular Bayesian priors for spatial smoothing. These are the Ising prior and two priors based on latent Gaussian MRFs. Concern is given to the selection of a suitable Markov chain Monte Carlo (MCMC) sampling scheme for each prior. The properties of the three priors and sampling schemes are investigated in the context of three empirical examples. The first is a simulated dataset, the second involves a confocal fluorescence microscopy dataset, while the third is based on the analysis of functional magnetic resonance imaging (fMRI) data. In the case of the Ising prior, single site and multi-site Swendsen-Wang sampling schemes are both considered. The single site scheme is shown to work consistently well, while it is shown that the Swendsen-Wang algorithm can have convergence problems. The sampling schemes for the priors are extended to generate the smoothing parameters, so that estimation becomes fully automatic. Although this works well, it is found that for highly contiguous images fixing smoothing parameters to very high values can improve results by injecting additional prior information concerning the level of contiguity in the image. The relative properties of the three binary MRFs are investigated, and it is shown how the Ising prior in particular defines sharp edges and encourages clustering. In addition, one of the latent Gaussian MRF priors is shown to be unable to distinguish between higher levels of smoothing. In the context of the fMRI example we also undertake a simulation study.     

Interval censored sampling plans for the gamma lifetime model

In this article, an acceptance sampling procedure for the gamma lifetime model is established under the interval censored test. Both producer and consumer risks are considered to develop the ordinary and approximate sampling plans with grouped data. Some of the proposed sampling plans are tabulated and the use of the tables is illustrated by an example. A sensitivity study is conducted to evaluate the influence of the length of the time interval on the proposed sampling plans.     

Design of progressively censored group sampling plans for Weibull distributions: An optimization problem

Optimization algorithms provides efficient solutions to many statistical problems. Essentially, the design of sampling plans for lot acceptance purposes is an optimization problem with several constraints, usually related to the quality levels required by the producer and the consumer. An optimal acceptance sampling plan is developed in this paper for the Weibull distribution with unknown scale parameter. The proposed plan combines grouping of items, sudden death testing in each group and progressive group removals, and its decision criterion is based on the uniformly most powerful life test. A mixed integer programming problem is first solved for determining the minimum number of failures required and the corresponding acceptance constant. The optimal number of groups is then obtained by minimizing a balanced estimation of the expected test cost. Excellent approximately optimal solutions are also provided in closed-forms. The sampling plan is considerably flexible and allows to save experimental time and cost. In general, our methodology achieves solutions that are quite robust to small variations in the Weibull shape parameter. A numerical example about a manufacturing process of gyroscopes is included for illustration.     

Arc refractor methods for adaptive importance sampling on large Bayesian networks under evidential reasoning

Approximate Bayesian inference by importance sampling derives probabilistic statements from a Bayesian network, an essential part of evidential reasoning with the network and an important aspect of many Bayesian methods. A critical problem in importance sampling on Bayesian networks is the selection of a good importance function to sample a network’s prior and posterior probability distribution. The initially optimal importance functions eventually start deviating from the optimal function when sampling a network’s posterior distribution given evidence, even when adaptive methods are used that adjust an importance function to the evidence by learning. In this article we propose a new family of Refractor Importance Sampling (RIS) algorithms for adaptive importance sampling under evidential reasoning. RIS applies “arc refractors” to a Bayesian network by adding new arcs and refining the conditional probability tables. The goal of RIS is to optimize the importance function for the posterior distribution and reduce the error variance of sampling. Our experimental results show a significant improvement of RIS over state-of-the-art adaptive importance sampling algorithms.     

Optimal lot disposition from Poisson–Lindley count data

Optimal sampling plans based on overdispersed defect counts for screening lots of outgoing and incoming goods are derived by minimizing the required sample size. Best inspection schemes provide appropriate protections to customers and manufacturers. The stochastic distribution of the number of defects per sampled unit is described by Poisson–Lindley models. Optimal frequentist and Bayesian decision rules for lot disposition are found by solving mixed integer nonlinear programming problems through simulation. The suggested criteria are based on likelihood and posterior odds ratios. The asymptotic normality of the quality score statistic is used to deduce explicit and reasonably accurate approximations of the optimal acceptance sampling plans. The Bayesian approach allows the practitioners to reduce the needed sample size for sentencing lots of high-quality products. For illustrative purposes, the proposed methods are applied to the manufacturing of copper wire.     

Hedging mortality/longevity risks of insurance portfolios for life insurer/annuity provider and financial intermediary

In this paper, we propose two risk hedge schemes in which a life insurer (an annuity provider) can transfer mortality (longevity) risk of a portfolio of life (annuity) exposures to a financial intermediary by paying the hedging premium of a mortality-linked security. The optimal units of the mortality-linked security which maximize hedge effectiveness for a life insurer (an annuity provider) can be derived as closed-form formulas under the risk hedge schemes. Numerical illustrations show that the risk hedge schemes can significantly hedge the downside risk of loss due to mortality (longevity) risk for the life insurer (annuity provider) under some stochastic mortality models. Besides, finding an optimal weight of a portfolio of life and annuity business, the financial intermediary can reduce the sensitivity to mortality rates but the model risk; a security loading may be imposed on the hedge premium for a higher probability of gain to compensate the financial intermediary for the inevitable model risk.