我国政府统计中多层次推断问题讨论

当前,多层次推断问题已成为我国在政府统计中推广抽样调查所面临的最大难题。本文回顾了现有解决多层次推断问题的方法,指出这些方法在使用上的限制;提出从改进估计的角度解决多层次推断问题,讨论了借助辅助信息改进估计的方法和建立统计模型进行推估的方法,并指出各种方法的优劣和适用情况;简要探讨了大数据背景下对解决多层次推断问题的一些启示。     

Storage Life Prediction for a High-Performance Capacitor Using Multi-Phase Wiener Degradation Model

Capacitance is a critical performance characteristic of high-voltage-pulse capacitor which is used to store and discharge electrical energy rapidly. The capacitors usually are stored for a long period of time before put into use. Experimental result and engineering experience indicate that the capacitance increases with storage time and will eventually exceed the failure threshold, which means that the capacitor may fail during storage. This is a typical mode of degradation failure for long storage products. Further, the capacitance degradation path can be extrapolated in several stages based on the shifting characteristics. That is, the capacitance increases slowly or fluctuates in the initial storage stage that lasts about three months. Then it increases sharply in the middle stage which lasts about four months. After the two stages, the capacitor enters into the third stage in which capacitance increases constantly. This degradation phenomenon motivates us to study the storage life prediction method based on multi-phase degradation path model. The storage performance degradation mechanism of high-voltage-pulse capacitor was investigated, which provides the physical basis for multi-phase Wiener degradation model. Identification procedure for the transition points in the degradation path was proposed using maximum likelihood principle (MLP). The result of Kruskal-Wallis test which is the method to test whether two populations are consistent or not in statistics showed that the transition points are statistically effective. Other parameters in the multi-phase degradation model are estimated with maximum likelihood estimation (MLE) after the transition points have been specified. The multi-phase Inverse Gaussian (IG) distribution for storage life was deduced for the capacitor, and the point and interval estimation procedure for reliable storage life are constructed with bootstrap method. The efficiency and effectiveness of the proposed multi-phase degradation model is compared with storage life prediction under single-phase condition.     

Testing non-nested log-linear models with pseudo estimator

As the number of random variables for the categorical data increases, the possible number of log-linear models which can be fitted to the data increases rapidly, so that various model selection methods are developed. However, we often found that some models chosen by different selection criteria do not coincide. In this paper, we propose a comparison method to test the final models which are non-nested. The statistic of Cox (1961, 1962) is applied to log-linear models for testing non-nested models, and the Kullback-Leibler measure of closeness (Pesaran 1987) is explored. In log-linear models, pseudo estimators for the expectation and the variance of Cox's statistic are not only derived but also shown to be consistent estimators.     

A strong law of large number for negatively dependent and non identical distributed random variables in the framework of sublinear expectation

Abstract

In this article, in the framework of sublinear expectation initiated by Peng, we derive a strong law of large numbers (SLLN) for negatively dependent and non identical distributed random variables. This result includes and extends some existing results. Furthermore, we give two examples of our result for applications.     

Additive Transformation Models for Recurrent Events

In this article, we propose a class of additive transformation models for recurrent event data, which includes the additive rates model as a special case. The new models offer great flexibility in formulating the effects of covariates on the mean function of recurrent events. Estimating equation approaches are developed for the model parameters, and asymptotic properties of the resulting estimators are established. In addition, a model checking procedure is presented to assess the adequacy of the model. The finite sample performance of the proposed estimators is examined through simulation studies, and an application to a bladder cancer study is presented.     

Empirical Analysis of Interval-Censored Failure Time Data with Linear Transformation Models

Interval-censored data naturally arise in many studies. For their regression analysis, many approaches have been proposed under various models and for most of them, the inference is carried out based on the asymptotic normality. In particular, Zhang et al. (2005) discussed the procedure under the linear transformation model. It is well-known that the symmetric property implied by the normal distribution may not be appropriate sometimes. Also the method could underestimate the variance of estimated parameters. This paper proposes an empirical likelihood-based procedure for the problem. Simulation and the analysis of a real data set are conducted to assess the performance of the procedure.     

A Multiple Imputation Approach to the Analysis of Current Status Data with the Additive Hazards Model

This article discusses regression analysis of current status data, which occur in many fields including cross-sectional studies, demographical investigations, and tumorigenicity experiments (Keiding, 1991 Keiding , N. ( 1991 ). Age-specific incidence and prevalence: a statistical perspective (with discussion) . J. Roy. Statist. Soc. Ser. A 154 : 371 – 412 .[Crossref] , [Google Scholar]; Sun 2006 Sun , J. ( 2006 ). The Statistical Analysis of Interval-Censored Failure Time Data . New York : Springer-Verlag . [Google Scholar]). For the problem, we focus on the situation where the survival time of interest can be described by the additive hazards model and a multiple imputation approach is presented for inference. A major advantage of the approach is its simplicity and it can be easily implemented by using the existing software packages for right-censored failure time data. Extensive simulation studies are conducted and indicate that the approach performs well for practical situations and is comparable to the existing methods. The methodology is applied to a set of current status data arising from a tumorigenicity experiment and the model checking is discussed.     

Nonparametric Tests for Stratum Effects in the Cox Model

Thispaper considers the stratified proportional hazards model witha focus on the assessment of stratum effects. The assessmentof such effects is often of interest, for example, in clinicaltrials. In this case, two relevant tests are the test of stratuminteraction with covariates and the test of stratum interactionwith baseline hazard functions. For the test of stratum interactionwith covariates, one can use the partial likelihood method (Kalbfleischand Prentice, 1980; Lin, 1994). For the test of stratum interactionwith baseline hazard functions, however, there seems to be noformal test available. We consider this problem and propose aclass of nonparametric tests. The asymptotic distributions ofthe tests are derived using the martingale theory. The proposedtests can also be used for survival comparisons which need tobe adjusted for covariate effects. The method is illustratedwith data from a lung cancer clinical trial.     

Designing and integrating composite networks for monitoring multivariate gaussian pollution fields

Networks of ambient monitoring stations are used to monitor environmental pollution fields such as those for acid rain and air pollution. Such stations provide regular measurements of pollutant concentrations. The networks are established for a variety of purposes at various times so often several stations measuring different subsets of pollutant concentrations can be found in compact geographical regions. The problem of statistically combining these disparate information sources into a single 'network' then arises. Capitalizing on the efficiencies so achieved can then lead to the secondary problem of extending this network. The subject of this paper is a set of 31 air pollution monitoring stations in southern Ontario. Each of these regularly measures a particular subset of ionic sulphate, sulphite, nitrite and ozone. However, this subset varies from station to station. For example only two stations measure all four. Some measure just one. We describe a Bayesian framework for integrating the measurements of these stations to yield a spatial predictive distribution for unmonitored sites and unmeasured concentrations at existing stations. Furthermore we show how this network can be extended by using an entropy maximization criterion. The methods assume that the multivariate response field being measured has a joint Gaussian distribution conditional on its mean and covariance function. A conjugate prior is used for these parameters, some of its hyperparameters being fitted empirically.