Canadian Journal of Statistics-Revue Canadienne De Statistique最新文献

Interval-censored data arise when a failure process is under intermittent observation and failure status is only known at assessment times. We consider the development of predictive algorithms when training samples involve interval censoring. Using censoring unbiased transformations and pseudo-observations, we define observed data loss functions, which are unbiased estimates of the corresponding complete data loss functions. We show that regression trees based on these loss functions can recover the tree structure and yield good predictive accuracy. An application is given to a study involving individuals with psoriatic arthritis where the aim is to identify genetic markers useful for the prediction of axial disease within 10 years of a baseline assessment.

Consider the problem of benchmarking small-area estimates under multiplicative models with positive parameters. The goal is to propose a loss function that guarantees positive constrained estimates of small-area parameters in this situation. The weighted precautionary loss function is introduced to solve the problem. Compared with the weighted Kullback–Leibler (KL) loss function, our proposed loss function penalizes underestimation of the small-area parameters of interest more for small values of parameters. This property is appealing when we estimate disease rates. It tends to give larger estimates of small-area parameters compared with those obtained under the KL loss function. The hierarchical empirical Bayes and constrained hierarchical empirical Bayes estimates of small-area parameters and their corresponding risk functions under the new proposed loss function are obtained. The performance of the proposed methods is investigated using simulation studies and a real dataset.

Data collected over time are common in applications and may contain censored or missing observations, making it difficult to use standard statistical procedures. This article proposes an algorithm to estimate the parameters of a censored linear regression model with errors serially correlated and innovations following a Student-$��t�$ distribution. This distribution is widely used in the statistical modelling of data containing outliers because its longer-than-normal tails provide a robust approach to handling such data. The maximum likelihood estimates of the proposed model are obtained through a stochastic approximation of the EM algorithm. The methods are applied to an environmental dataset regarding ammonia-nitrogen concentration, which is subject to a limit of detection (left censoring) and contains missing observations. Additionally, two simulation studies are conducted to examine the asymptotic properties of the estimates and the robustness of the model. The proposed algorithm and methods are implemented in the R package ARCensReg.

We propose a new functional additive hazards model to investigate the potential effects of functional and scalar predictors on mortality risks, and develop a penalized least squares estimation method for model parameters based on a pseudoscore estimating equation. A reproducing kernel Hilbert space approach is used to establish the consistency, convergence rate, and joint asymptotic distribution of the resulting estimators for finite-dimensional and infinite-dimensional parameters. Our simulation studies demonstrate that the proposed estimation procedure performs well. For illustration, we apply the proposed method to the Medical Information Mart for Intensive Care III dataset.

We propose a Bayesian nonparametric clustering approach to study the spatial heterogeneity effect for functional data observed at spatially correlated locations. We consider a geographically weighted Chinese restaurant process equipped with a conditional autoregressive prior to capture fully the spatial correlation of function curves. To sample efficiently from our model, we customize a prior called Quadratic Gamma, which ensures conjugacy. We design a Markov chain Monte Carlo algorithm to infer simultaneously the posterior distributions of the number of groups and the grouping configurations. The superior numerical performance of the proposed method over competing methods is demonstrated using simulated examples and a U.S. annual precipitation study.

We consider data integration problems where correlated data are collected from multiple platforms. Within each platform, there are linear relationships between the responses and a collection of predictors. We extend the linear models to include random errors coming from a much wider family of sub-Gaussian and subexponential distributions. The goal is to select important predictors across multiple platforms, where the number of predictors and the number of observations both increase to infinity. We combine the marginal densities of the responses obtained from different platforms to form a composite likelihood and propose a model selection criterion based on Bayesian composite posterior probabilities. Under some regularity conditions, we prove that the model selection criterion is consistent to recover the union support of the predictors with divergent true model size.

Semicontinuous data frequently occur in longitudinal studies. The popular two-part modelling approach deals with longitudinal semicontinuous data by analyzing the occurrence of positive values and the intensity of positive values separately; however, this separation may break down the natural sequence of semicontinuous data within a subject and destroy its serial dependence structure. In this article, we introduce a Tweedie compound Poisson mixed model to study the occurrence of positive values and the quantity of the semicontinuous response simultaneously. In our approach, covariate effects on the semicontinuous response are assessed directly. The correlation within a subject and the unobserved heterogeneity are incorporated with serially correlated nonparametric random effects. Our model unifies subject-specific and population-averaged interpretations. We illustrate the approach with applications to a Brief Symptom Inventory study and an infants' fluoride intake study.

Item nonresponse is a common issue in surveys. Because unadjusted estimators may be biased in the presence of nonresponse, it is common practice to impute the missing values with the objective of reducing the nonresponse bias as much as possible. However, commonly used imputation procedures may lead to unstable estimators of population totals/means when influential units are present in the set of respondents. In this article, we consider the class of multiply robust imputation procedures that provide some protection against the failure of underlying model assumptions. We develop an efficient version of multiply robust estimators based on the concept of conditional bias, a measure of influence. We present the results of a simulation study to show the benefits of our proposed method in terms of bias and efficiency.

This article considers the challenge of designing football group draw mechanisms, which have a uniform distribution over all valid draw assignments, but are also entertaining, practical and transparent. Although this problem is trivial in completely symmetric problems, it becomes challenging when there are draw constraints that are not exchangeable across each of the competing teams, so that symmetry breaks down. We explain how to simulate the FIFA sequential draw method and compute the nonuniformity of its draws by comparison with a uniform rejection sampler. We then propose several practical methods of achieving the uniform distribution while still using balls and bowls in a way which is suitable for a televised draw. The solutions can also be carried out interactively. The general methodology we provide can readily be transported to different competition draws and is not restricted to football events.