日期精度：到日 2256 篇 · 到月 2354 篇 · 到年 2207 篇
自助法鞍点近似
We compare saddlepoint approximations to the exact distributions of a studentized mean and to its bootstrap approximation. We show that, on bounded sets, these empirical saddlepoint approximations achieve second order relative errors uniformly. We also consider the relative errors for larger deviations. It follows that the studentized-t bootstrap p-value and the coverage of the bootstrap confidence interval have second order relative errors.
基于估计方程的人类家系遗传分离分析
Studies in genetic epidemiology are generally required to collect phenotypes on subjects from human pedigrees that consist of blood relatives and their spouses. One of the main objectives in analyzing pedigree data is to study the segregation of underlying genes that may predispose certain phenotypes and thus cause phenotypic aggregation within pedigrees. Most classical segregation methods are maximum likelihood-based under highly specified distributional assumptions for…
抽样调查精度估计：方差估计量的选择准则
We advocate the 'mean squared error of predicted squared error' as a universal criterion for the choice between variance estimators in sample survey inference. The predictive nature of the measure rewards variance estimators adapted to the amount of information in the actual sample. This makes the new measure more satisfactory than the simpler mean squared error of the variance estimator. The criterion turns out to be the same for design-based as for model-based inference,…
扩散过程离散观测的贝叶斯因子
We present an approach to model selection for a time series of data on a fine time scale. The underlying process generating the data is modelled as a continuous time stochastic process. The underlying continuous processes are assumed to be diffusions with time varying drift and diffusion coefficient. Several approaches to modelling the diffusion coefficient are described. To perform model selection, we propose an approximation to the Bayes factor that uses only the discrete…
生物等效性检验对应的置信区间
Previously, the connection between bioequivalence tests and their associated confidence intervals was not well understood. In fact, the α-level two one-sided tests approach was thought to be associated with a 1−2α confidence interval. In this paper, we build up connections which allow us to construct 1−α confidence intervals from α-level tests and vice versa. When applied to the α-level two one-sided tests, the resultant 1−-α confidence intervals are properly contained in…
高阶渐近与似然原则：单参数模型
Inferences based on higher-order asymptotics, in contrast to those based on standard first-order methods, approximate exact inferences closely enough to depend on the reference set. We use this feature to quantify the effect of the reference set on exact inferences. In particular, the difference in P-values for two reference sets can be at most O(n−½), and it is seen that one can often identify the coefficient of n−½. Although this is useful from a foundational viewpoint,…
比例风险模型下生存曲线的置信带
In this paper, we show how to construct simultaneous confidence bands for the subject-specific survival curve under the Cox proportional hazards model. The idea is to approximate the distribution of the normalized cumulative hazard estimator by a zero-mean Gaussian process whose distribution can be easily generated through simulation. Numerical studies indicate that the proposed bands are appropriate for practical use. A liver disease example is presented.
加性风险模型的半参数分析
In contrast to the proportional hazards model, the additive risk model specifies that the hazard function associated with a set of possibly time-varying covariates is the sum of, rather than the product of, the baseline hazard function and the regression function of covariates. This formulation describes a different aspect of the association between covariates and the failure time than the proportional hazards model, and is more plausible than the latter for many…
Gibbs 抽样器的协方差结构：估计量与增广方案比较
We study the covariance structure of a Markov chain generated by the Gibbs sampler, with emphasis on data augmentation. When applied to a Bayesian missing data problem, the Gibbs sampler produces two natural approximations for the posterior distribution of the parameter vector: the empirical distribution based on the sampled values of the parameter vector, and a mixture of complete data posteriors. We prove that Rao-Blackwellization causes a one-lag delay for the…
随机向量差分方程渐近性质的一则札记
This note analyses a necessary condition for asymptotic normality of the maximum likelihood estimator in a stationary stochastic vector difference equation. It is shown that this condition is satisfied if the error variance matrix is positive definite.
非正则回归
In an earlier paper, the author has studied asymptotic properties of maximum likelihood estimates for a class of nonregular models in which the range of the distribution depends on unknown parameters. Examples include Weibull and extreme value distributions. The present paper is concerned with the extension of this theory to the case when covariates are present. The proposed solution involves solving a linear programming problem for the regression parameters, followed by…
移动平均模型的一种简单非迭代估计量
We examine a simple noniterative estimator for the parameters of a general moving average process. This non-maximum-likelihood estimator derives moving average model parameters directly from the coefficients of an approximating autoregressive model. The estimator is evaluated through asymptotic expansions and by simulation, and is also compared with maximum likelihood and the related estimator of Durbin (1959). The comparison with maximum likelihood by simulation suggests a…
非平稳状态空间模型的快速似然计算与预测
A recursive procedure for initializing the Kalman filter is displayed. The recursion is for nonstationary state space models. The procedure imposes small computational and programming burden over and above the Kalman filter. The procedure is superior to other suggested approaches in both computational speed and general applicability. General properties of the method are investigated. Details of the initialization for the ARIMA (p,d,q) and basic structural models are…
非正则回归
In an earlier paper, the author has studied asymptotic properties of maximum likelihood estimates for a class of nonregular models in which the range of the distribution depends on unknown parameters. Examples include Weibull and extreme value distributions. The present paper is concerned with the extension of this theory to the case when covariates are present. The proposed solution involves solving a linear programming problem for the regression parameters, followed by…
抽样调查推断中的精度估计：方差估计量选择准则
We advocate the 'mean squared error of predicted squared error' as a universal criterion for the choice between variance estimators in sample survey inference. The predictive nature of the measure rewards variance estimators adapted to the amount of information in the actual sample. This makes the new measure more satisfactory than the simpler mean squared error of the variance estimator. The criterion turns out to be the same for design-based as for model-based inference,…
与生物等效性检验相关的置信区间
Previously, the connection between bioequivalence tests and their associated confidence intervals was not well understood. In fact, the α-level two one-sided tests approach was thought to be associated with a 1−2α confidence interval. In this paper, we build up connections which allow us to construct 1−α confidence intervals from α-level tests and vice versa. When applied to the α-level two one-sided tests, the resultant 1−-α confidence intervals are properly contained in…
分组生存数据边际似然的最大化
Grouped failure time data occur in studies where subjects are monitored periodically to determine whether failure has occurred in the intervening interval. Here the model under consideration is Cox's (1972, 1975) proportional hazards model, but the commonly used method of partial likelihood needs modification with grouped data due to a potentially large number of ties. This paper demonstrates how the Monte Carlo em algorithm (Wei & Tanner, 1990) can be used on grouped data…
Gibbs 抽样的协方差结构：估计量与数据增广方案的比较
We study the covariance structure of a Markov chain generated by the Gibbs sampler, with emphasis on data augmentation. When applied to a Bayesian missing data problem, the Gibbs sampler produces two natural approximations for the posterior distribution of the parameter vector: the empirical distribution based on the sampled values of the parameter vector, and a mixture of complete data posteriors. We prove that Rao-Blackwellization causes a one-lag delay for the…
自助法鞍点近似
We compare saddlepoint approximations to the exact distributions of a studentized mean and to its bootstrap approximation. We show that, on bounded sets, these empirical saddlepoint approximations achieve second order relative errors uniformly. We also consider the relative errors for larger deviations. It follows that the studentized-t bootstrap p-value and the coverage of the bootstrap confidence interval have second order relative errors.
移动平均模型的简单非迭代估计量
We examine a simple noniterative estimator for the parameters of a general moving average process. This non-maximum-likelihood estimator derives moving average model parameters directly from the coefficients of an approximating autoregressive model. The estimator is evaluated through asymptotic expansions and by simulation, and is also compared with maximum likelihood and the related estimator of Durbin (1959). The comparison with maximum likelihood by simulation suggests a…