Robust estimation of the correlation matrix of longitudinal data

We propose a double-robust procedure for modeling the correlation matrix of a longitudinal dataset. It is based on an alternative Cholesky decomposition of the form Σ=DLL ^? D where D is a diagonal matrix proportional to the square roots of the diagonal entries of Σ and L is a unit lower-triangular matrix determining solely the correlation matrix. The first robustness is with respect to model misspecification for the innovation variances in D, and the second is robustness to outliers in the data. The latter is handled using heavy-tailed multivariate t-distributions with unknown degrees of freedom. We develop a Fisher scoring algorithm for computing the maximum likelihood estimator of the parameters when the nonredundant and unconstrained entries of (L,D) are modeled parsimoniously using covariates. We compare our results with those based on the modified Cholesky decomposition of the form LD ² L ^? using simulations and a real dataset.     

Maximum modulus confidence bands

A family of confidence bands (simultaneous confidence regions) is given for EY = x′β that are piecewise-linear in x. Normality is assumed. These confidence bands are advocated over the usual hyperbolic band when the region of prime interest is distant from ${\overline{\bf x}}$ . In particular, this is the case when x?=?x(t) for time t and future time is of primary interest, that is for the prediction problem. For the case x′?=?(1, t), the family of bands includes that of Graybill and Bowden (J Am Stat Assoc 62:403–408, 1967).     

Robust model-based sampling designs

We investigate methods for the design of sample surveys, and address the traditional resistance of survey samplers to the use of model-based methods by incorporating model robustness at the design stage. The designs are intended to be sufficiently flexible and robust that resulting estimates, based on the designer’s best guess at an appropriate model, remain reasonably accurate in a neighbourhood of this central model. Thus, consider a finite population of N units in which a survey variable Y is related to a q dimensional auxiliary variable x. We assume that the values of x are known for all N population units, and that we will select a sample of n≤N population units and then observe the n corresponding values of Y. The objective is to predict the population total $T=\sum_{i=1}^{N}Y_{i}$ . The design problem which we consider is to specify a selection rule, using only the values of the auxiliary variable, to select the n units for the sample so that the predictor has optimal robustness properties. We suppose that T will be predicted by methods based on a linear relationship between Y—possibly transformed—and given functions of x. We maximise the mean squared error of the prediction of T over realistic neighbourhoods of the fitted linear relationship, and of the assumed variance and correlation structures. This maximised mean squared error is then minimised over the class of possible samples, yielding an optimally robust (‘minimax’) design. To carry out the minimisation step we introduce a genetic algorithm and discuss its tuning for maximal efficiency.     

Monitoring the conforming fraction of high-quality processes using a control chart p under a small sample size and an alternative estimator

In this paper, we propose to change the traditional monitored statistic in a control chart p, by changing the sampling proportion ${\hat{p}}$ to a new statistics denoted as ${\tilde{p}}$ . We aim to minimize problems in designing the control chart p for high quality processes when only a small sample size is available. The idea of the new statistics is simple, as it involves taking two independent samples of a Bernoulli population. From each sample, the sampling proportion is calculated, and the new statistic to monitor is the weighted mean of the sampling proportion of each sample employed to weight the overall sampling proportion. We note that the control chart p that employs the new ${\tilde{p}}$ statistic provides more in-control values of average run length closer to the usual fixed value of 370 than the traditional statistic, that is, the sampling proportion. Numerical examples illustrate the new proposal.     

Bayesian variable selection for correlated covariates via colored cliques

We propose a Bayesian method to select groups of correlated explanatory variables in a linear regression framework. We do this by introducing in the prior distribution assigned to the regression coefficients a random matrix $G$ that encodes the group structure. The groups can thus be inferred by sampling from the posterior distribution of $G$ . We then give a graph-theoretic interpretation of this random matrix $G$ as the adjacency matrix of cliques. We discuss the extension of the groups from cliques to more general random graphs, so that the proposed approach can be viewed as a method to find networks of correlated covariates that are associated with the response.     

L models and multiple regressions designs

Given an orthogonal model

${{\bf \lambda}}=\sum_{i=1}^m{{{\bf X}}_i}{\boldsymbol{\alpha}}_i$

an L model

${{\bf y}}={\bf L}\left(\sum_{i=1}^m{{{\bf X}}_i}{\boldsymbol{\alpha}}_i\right)+{\bf e}$

is obtained, and the only restriction is the linear independency of the column vectors of matrix L. Special cases of the L models correspond to blockwise diagonal matrices L = D(L ₁, . . . , L _c ). In multiple regression designs this matrix will be of the form

${\bf L}={\bf D}(\check{{\bf X}}_1,\ldots,\check{{\bf X}}_{c})$

with \({\check{{\bf X}}_j, j=1,\ldots,c}\) the model matrices of the individual regressions, while the original model will have fixed effects. In this way, we overcome the usual restriction of requiring all regressions to have the same model matrix.

     

On exact and optimal single-sampling plans by variables

We deal with sampling by variables with two-way protection in the case of a $N\>(\mu ,\sigma ^2)$ distributed characteristic with unknown $\sigma $ . The LR sampling plan proposed by Lieberman and Resnikoff (JASA 50: 457 ${-}$ 516, 1955) and the BSK sampling plan proposed by Bruhn-Suhr and Krumbholz (Stat. Papers 31: 195–207, 1990) are based on the UMVU and the plug-in estimator, respectively. For given $p_1$ (AQL), $p_2$ (RQL) and $\alpha ,\beta $ (type I and II errors) we present an algorithm allowing to determine the optimal LR and BSK plans having minimal sample size among all plans satisfying the corresponding two-point condition on the OC. An R (R: A language and environment for statistical computing, R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org/ 2012) package, ExLiebeRes‘ (Krumbholz and Steuer ExLiebeRes: calculating exact LR- and BSK-plans, R-package version 0.9.9. http://exlieberes.r-forge.r-project.org 2012) implementing that algorithm is provided to the public.     

Efficient Bayesian estimation of the multivariate Double Chain Markov Model

The Double Chain Markov Model (DCMM) is used to model an observable process $Y = \{Y_{t}\}_{t=1}^{T}$ as a Markov chain with transition matrix, $P_{x_{t}}$ , dependent on the value of an unobservable (hidden) Markov chain $\{X_{t}\}_{t=1}^{T}$ . We present and justify an efficient algorithm for sampling from the posterior distribution associated with the DCMM, when the observable process Y consists of independent vectors of (possibly) different lengths. Convergence of the Gibbs sampler, used to simulate the posterior density, is improved by adding a random permutation step. Simulation studies are included to illustrate the method. The problem that motivated our model is presented at the end. It is an application to real data, consisting of the credit rating dynamics of a portfolio of financial companies where the (unobserved) hidden process is the state of the broader economy.     

Estimation in a competing risks proportional hazards model under length-biased sampling with censoring

What population does the sample represent? The answer to this question is of crucial importance when estimating a survivor function in duration studies. As is well-known, in a stationary population, survival data obtained from a cross-sectional sample taken from the population at time $t_0$ represents not the target density $f(t)$ but its length-biased version proportional to $tf(t)$ , for $t>0$ . The problem of estimating survivor function from such length-biased samples becomes more complex, and interesting, in presence of competing risks and censoring. This paper lays out a sampling scheme related to a mixed Poisson process and develops nonparametric estimators of the survivor function of the target population assuming that the two independent competing risks have proportional hazards. Two cases are considered: with and without independent censoring before length biased sampling. In each case, the weak convergence of the process generated by the proposed estimator is proved. A well-known study of the duration in power for political leaders is used to illustrate our results. Finally, a simulation study is carried out in order to assess the finite sample behaviour of our estimators.     

Linear models that allow perfect estimation

The general Gauss–Markov model, Y = Xβ + e, E(e) = 0, Cov(e) = σ ² V, has been intensively studied and widely used. Most studies consider covariance matrices V that are nonsingular but we focus on the most difficult case wherein C(X), the column space of X, is not contained in C(V). This forces V to be singular. Under this condition there exist nontrivial linear functions of Q′Xβ that are known with probability 1 (perfectly) where ${C(Q)=C(V)^\perp}$ . To treat ${C(X) \not \subset C(V)}$ , much of the existing literature obtains estimates and tests by replacing V with a pseudo-covariance matrix T = V + XUX′ for some nonnegative definite U such that ${C(X) \subset C(T)}$ , see Christensen (Plane answers to complex questions: the theory of linear models, 2002, Chap. 10). We find it more intuitive to first eliminate what is known about Xβ and then to adjust X while keeping V unchanged. We show that we can decompose β into the sum of two orthogonal parts, β = β ₀ + β ₁, where β ₀ is known. We also show that the unknown component of X β is ${X\beta_1 \equiv \tilde{X} \gamma}$ , where ${C(\tilde{X})=C(X)\cap C(V)}$ . We replace the original model with ${Y-X\beta_0=\tilde{X}\gamma+e}$ , E(e) = 0, ${Cov(e)=\sigma^2V}$ and perform estimation and tests under this new model for which the simplifying assumption ${C(\tilde{X}) \subset C(V)}$ holds. This allows us to focus on the part of that parameters that are not known perfectly. We show that this method provides the usual estimates and tests.     

On the joint distribution of success runs of several lengths in the sequence of MBT and its applications

Let \(X_1 ,X_2 ,\ldots ,X_n \) be a sequence of Markov Bernoulli trials (MBT) and \(\underline{X}_n =( {X_{n,k_1 } ,X_{n,k_2 } ,\ldots ,X_{n,k_r } })\) be a random vector where \(X_{n,k_i } \) represents the number of occurrences of success runs of length \(k_i \,( {i=1,2,\ldots ,r})\) . In this paper the joint distribution of \(\underline{X}_n \) in the sequence of \(n\) MBT is studied using method of conditional probability generating functions. Five different counting schemes of runs namely non-overlapping runs, runs of length at least \(k\) , overlapping runs, runs of exact length \(k\) and \(\ell \) -overlapping runs (i.e. \(\ell \) -overlapping counting scheme), \(0\le \ell are considered. The pgf of joint distribution of \(\underline{X}_n \) is obtained in terms of matrix polynomial and an algorithm is developed to get exact probability distribution. Numerical results are included to demonstrate the computational flexibility of the developed results. Various applications of the joint distribution of \(\underline{X}_n \) such as in evaluation of the reliability of \(( {n,f,k})\!\!:\!\!G\) and \(\!:\!\!G\) system, in evaluation of quantities related to start-up demonstration tests, acceptance sampling plans are also discussed.     

On conditional variance estimation in nonparametric regression

In this paper we consider a nonparametric regression model in which the conditional variance function is assumed to vary smoothly with the predictor. We offer an easily implemented and fully Bayesian approach that involves the Markov chain Monte Carlo sampling of standard distributions. This method is based on a technique utilized by Kim, Shephard, and Chib (in Rev. Econ. Stud. 65:361–393, 1998) for the stochastic volatility model. Although the (parametric or nonparametric) heteroscedastic regression and stochastic volatility models are quite different, they share the same structure as far as the estimation of the conditional variance function is concerned, a point that has been previously overlooked. Our method can be employed in the frequentist context and in Bayesian models more general than those considered in this paper. Illustrations of the method are provided.     

Conditional MSE-based discrimination of the sample mean and the post-stratification estimator in population sampling

Whenever a random sample is drawn from a stratified population, the post-stratification estimator $\tilde X$ usually is preferred to the sample mean $\tilde X$ , when the population mean is to be estimated. This is due to the fact that the variance of $\tilde X$ is asymptotically smaller than that of $\tilde X$ , while both estimators are asymptotically unbiased. However, this only holds looking at post-stratification unconditionally, when strata sample sizes are random. Conditioned on the realized sample sizes, the MSE of $\tilde X$ can be higher than that of $\tilde X$ which means that $\tilde X$ should be preferred to $\tilde X$ , even if it is biased. The conditional MSE difference of $\tilde X$ and $\tilde X$ is estimated, and using this estimation and its variance a heuristic test based on the Vysochanskiî-Petunin inequality is derived.     

Covariance matrix of the bias-corrected maximum likelihood estimator in generalized linear models

For the first time, we obtain a general formula for the \(n^{-2}\) asymptotic covariance matrix of the bias-corrected maximum likelihood estimators of the linear parameters in generalized linear models, where \(n\) is the sample size. The usefulness of the formula is illustrated in order to obtain a better estimate of the covariance of the maximum likelihood estimators and to construct better Wald statistics. Simulation studies and an application support our theoretical results.     

On Bahadur representation for sample quantiles under α-mixing sequence

In this paper, by relaxing the mixing coefficients to α(n) = O(n ^?β), β > 3, we investigate the Bahadur representation of sample quantiles under α-mixing sequence and obtain the rate as ${O(n^{-\frac{1}{2}}(\log\log n\cdot\log n)^{\frac{1}{2}})}$ . Meanwhile, for any δ > 0, by strengthening the mixing coefficients to α(n) = O(n ^?β ), ${\beta > \max\{3+\frac{5}{1+\delta},1+\frac{2}{\delta}\}}$ , we have the rate as ${O(n^{-\frac{3}{4}+\frac{\delta}{4(2+\delta)}}(\log\log n\cdot \log n)^{\frac{1}{2}})}$ . Specifically, if ${\delta=\frac{\sqrt{41}-5}{4}}$ and ${\beta > \frac{\sqrt{41}+7}{2}}$ , then the rate is presented as ${O(n^{-\frac{\sqrt{41}+5}{16}}(\log\log n\cdot \log n)^{\frac{1}{2}})}$ .     

Consistency of a nonparametric conditional mode estimator for random fields

Given a stationary multidimensional spatial process $\left\{ Z_{\mathbf{i}}=\left( X_{\mathbf{i}},\ Y_{\mathbf{i}}\right) \in \mathbb R ^d\right. \left. \times \mathbb R ,\mathbf{i}\in \mathbb Z ^{N}\right\} $ , we investigate a kernel estimate of the spatial conditional mode function of the response variable $Y_{\mathbf{i}}$ given the explicative variable $X_{\mathbf{i}}$ . Consistency in $L^p$ norm and strong convergence of the kernel estimate are obtained when the sample considered is a $\alpha $ -mixing sequence. An application to real data is given in order to illustrate the behavior of our methodology.     

Importance tempering

Simulated tempering (ST) is an established Markov chain Monte Carlo (MCMC) method for sampling from a multimodal density π(θ). Typically, ST involves introducing an auxiliary variable k taking values in a finite subset of [0,1] and indexing a set of tempered distributions, say π _k (θ)∝ π(θ) ^k . In this case, small values of k encourage better mixing, but samples from π are only obtained when the joint chain for (θ,k) reaches k=1. However, the entire chain can be used to estimate expectations under π of functions of interest, provided that importance sampling (IS) weights are calculated. Unfortunately this method, which we call importance tempering (IT), can disappoint. This is partly because the most immediately obvious implementation is naïve and can lead to high variance estimators. We derive a new optimal method for combining multiple IS estimators and prove that the resulting estimator has a highly desirable property related to the notion of effective sample size. We briefly report on the success of the optimal combination in two modelling scenarios requiring reversible-jump MCMC, where the naïve approach fails.     

A non-iterative sampling Bayesian method for linear mixed models with normal independent distributions

In this paper, we utilize normal/independent (NI) distributions as a tool for robust modeling of linear mixed models (LMM) under a Bayesian paradigm. The purpose is to develop a non-iterative sampling method to obtain i.i.d. samples approximately from the observed posterior distribution by combining the inverse Bayes formulae, sampling/importance resampling and posterior mode estimates from the expectation maximization algorithm to LMMs with NI distributions, as suggested by Tan et al. [33 Tan, M., Tian, G. and Ng, K. 2003. A noniterative sampling method for computing posteriors in the structure of EM-type algorithms. Statist. Sinica, 13(3): 625–640. [Web of Science ®] , [Google Scholar]]. The proposed algorithm provides a novel alternative to perfect sampling and eliminates the convergence problems of Markov chain Monte Carlo methods. In order to examine the robust aspects of the NI class, against outlying and influential observations, we present a Bayesian case deletion influence diagnostics based on the Kullback–Leibler divergence. Further, some discussions on model selection criteria are given. The new methodologies are exemplified through a real data set, illustrating the usefulness of the proposed methodology.