Hybrid journal (It can contain Open Access articles) ISSN (Print) 0006-3444 - ISSN (Online) 1464-3510 Published by Oxford University Press[370 journals]

Authors:Gao X; Carroll RJ. Abstract: SummaryWe consider situations where the data consist of a number of responses for each individual, which may include a mix of discrete and continuous variables. The data also include a class of predictors, where the same predictor may have different physical measurements across different experiments depending on how the predictor is measured. The goal is to select which predictors affect any of the responses, where the number of such informative predictors tends to infinity as the sample size increases. There are marginal likelihoods for each experiment; we specify a pseudolikelihood combining the marginal likelihoods, and propose a pseudolikelihood information criterion. Under regularity conditions, we establish selection consistency for this criterion with unbounded true model size. The proposed method includes a Bayesian information criterion with appropriate penalty term as a special case. Simulations indicate that data integration can dramatically improve upon using only one data source. PubDate: 2017-05-09

Authors:Farewell DM; Huang CC, Didelez VV. Abstract: SummaryLikelihood factors that can be disregarded for inference are termed ignorable. We demonstrate that close ties exist between ignorability and identification of causal effects by covariate adjustment. A graphical condition, stability, plays a role analogous to that of missingness at random, but is applicable to general longitudinal data. Our formulation of ignorability does not depend on any notion of missing data, so is appealing in situations where missing data may not actually exist. Several examples illustrate how stability may be assessed. PubDate: 2017-05-08

Authors:Papaspiliopoulos OO; Rossell DD. Abstract: SummaryWe propose a scalable algorithmic framework for exact Bayesian variable selection and model averaging in linear models under the assumption that the Gram matrix is block-diagonal, and as a heuristic for exploring the model space for general designs. In block-diagonal designs our approach returns the most probable model of any given size without resorting to numerical integration. The algorithm also provides a novel and efficient solution to the frequentist best subset selection problem for block-diagonal designs. Posterior probabilities for any number of models are obtained by evaluating a single one-dimensional integral, and other quantities of interest such as variable inclusion probabilities and model-averaged regression estimates are obtained by an adaptive, deterministic one-dimensional numerical integration. The overall computational cost scales linearly with the number of blocks, which can be processed in parallel, and exponentially with the block size, rendering it most adequate in situations where predictors are organized in many moderately-sized blocks. For general designs, we approximate the Gram matrix by a block-diagonal matrix using spectral clustering and propose an iterative algorithm that capitalizes on the block-diagonal algorithms to explore efficiently the model space. All methods proposed in this paper are implemented in the R library mombf. PubDate: 2017-04-24

Authors:Wang J; Shen X, Sun Y, et al. Abstract: SummaryAutomatic tagging by key words and phrases is important in multi-label classification of a document. In this paper, we first introduce a tagging loss to measure the discrepancy between predicted and actual tag sets, which is expressed in terms of a sum of weighted pairwise margins between two tags by their degree of similarity. We then construct a regularized empirical loss to incorporate linguistic knowledge, and identify a tagger maximizing the separations between the pairwise margins. One salient feature of the proposed method is its capability to identify novel tags absent from a training sample by using their similarity to existing tags. Computationally, the proposed method is implemented by an alternating direction method of multipliers, integrated with a difference convex algorithm. This permits scalable computation. We show that the method achieves accurate tagging, and that it compares favourably with existing methods. Finally, we apply the proposed method to tagging a Reuters news dataset. PubDate: 2017-04-21

Authors:Lee S; Sun W, Wright FA, et al. Abstract: SummaryUnobserved environmental, demographic and technical factors canadversely affect the estimation and testing of the effects ofprimary variables. Surrogate variable analysis, proposed to tacklethis problem, has been widely used in genomic studies. To estimatehidden factors that are correlated with the primary variables,surrogate variable analysis performs principal component analysiseither on a subset of features or on all features, but weightingeach differently. However, existing approaches may fail to identifyhidden factors that are strongly correlated with the primaryvariables, and the extra step of feature selection and weightcalculation makes the theoretical investigation of surrogatevariable analysis challenging. In this paper, we propose an improvedsurrogate variable analysis, using all measured features, that has anatural connection with restricted least squares, which allows us tostudy its theoretical properties. Simulation studies and real-dataanalysis show that the method is competitive with state-of-the-artmethods. PubDate: 2017-04-21

Authors:Kong X. Abstract: SummaryIn this paper, we introduce a local principal component analysis approach to determining the number of common factors of a continuous-time factor model with time-varying factor loadings using high-frequency data. The model is approximated locally on shrinking blocks using discrete-time factor models. The number of common factors is estimated by minimizing the penalized aggregated mean squared residual error over all shrinking blocks. While the local mean squared residual error on each block converges at rate $\min(n^{1/4}, p)$, where $n$ is the sample size and $p$ is the dimension, the aggregated mean squared residual error converges at rate $\min(n^{1/2}, p)$; this achieves the convergence rate of the penalized criterion function of the global principal component analysis method, assuming restrictive constant factor loading. An estimator of the number of factors based on the local principal component analysis is consistent. Simulation results justify the performance of our estimator. A real financial dataset is analysed. PubDate: 2017-04-21

Authors:Sun F; Tang B. Abstract: SummaryOrthogonal Latin hypercubes provide a class of useful designs for computer experiments. Among the available methods for constructing such designs, the method of rotation is particularly prominent due to its theoretical appeal as well as its space-filling properties. This paper presents a general method of rotation for constructing orthogonal Latin hypercubes, making the rotation idea applicable to many more situations than the original method allows. In addition to general theoretical results, many new orthogonal Latin hypercubes are obtained and tabulated. PubDate: 2017-04-21

Authors:Lam C; Feng P, Hu C. Abstract: SummaryIntegrated covariance matrices arise in intraday models of asset returns, which allow volatility to change over the trading day. When the number of assets is large, the natural estimator of such a matrix suffers from bias due to extreme eigenvalues. We introduce a novel nonlinear shrinkage estimator for the integrated covariance matrix which shrinks the extreme eigenvalues of a realized covariance matrix back to an acceptable level, and enjoys a certain asymptotic efficiency when the number of assets is of the same order as the number of data points. Novel maximum exposure and actual risk bounds are derived when our estimator is used in constructing the minimum variance portfolio. In simulations and a real-data analysis, our estimator performs favourably in comparison with other methods. PubDate: 2017-04-21

Authors:Ding PP; Vanderweele TJ, Robins JM. Abstract: SummaryDrawing causal inference with observational studies is the central pillar of many disciplines. One sufficient condition for identifying the causal effect is that the treatment-outcome relationship is unconfounded conditional on the observed covariates. It is often believed that the more covariates we condition on, the more plausible this unconfoundedness assumption is. This belief has had a huge impact on practical causal inference, suggesting that we should adjust for all pretreatment covariates. However, when there is unmeasured confounding between the treatment and outcome, estimators adjusting for some pretreatment covariate might have greater bias than estimators that do not adjust for this covariate. This kind of covariate is called a bias amplifier, and includes instrumental variables that are independent of the confounder and affect the outcome only through the treatment. Previously, theoretical results for this phenomenon have been established only for linear models. We fill this gap in the literature by providing a general theory, showing that this phenomenon happens under a wide class of models satisfying certain monotonicity assumptions. PubDate: 2017-04-17

Authors:Linero AR. Abstract: SummaryIn longitudinal clinical trials, one often encounters missingness that is thought to be nonignorable. Such missingness introduces identifiability issues, resulting in causal effects being nonparametrically unidentified; it is then prudent to conduct a sensitivity analysis to assess how much of the inference is being driven by untestable assumptions needed to identify the effects of interest. We introduce a Bayesian nonparametric framework for conducting inference in the presence of nonignorable, nonmonotone missingness. This framework focuses on the specification of an auxiliary working prior on the space of complete data generating mechanisms. This prior induces a prior on the observed data generating mechanism, which is then used in conjunction with an identifying restriction to conduct inference. Advantages of this approach include a flexible modelling framework, access to simple computational methods, strong theoretical support, straightforward sensitivity analysis, and applicability to nonmonotone missingness. PubDate: 2017-04-17

Authors:Dalal O; Rajaratnam B. Abstract: SummarySeveral methods have recently been proposed for estimating sparse Gaussian graphical models using $\ell_{1}$-regularization on the inverse covariance or precision matrix. Despite recent advances, contemporary applications require even faster methods to handle ill-conditioned high-dimensional datasets. In this paper, we propose a new method for solving the sparse inverse covariance estimation problem using the alternating minimization algorithm, which effectively works as a proximal gradient algorithm on the dual problem. Our approach has several advantages: it is faster than state-of-the-art algorithms by many orders of magnitude; its global linear convergence has been rigorously demonstrated, underscoring its good theoretical properties; it facilitates additional constraints on pairwise or marginal relationships between feature pairs based on domain-specific knowledge; and it is better at handling extremely ill-conditioned problems. Our algorithm is shown to be more accurate and faster on simulated and real datasets. PubDate: 2017-04-17

Authors:Ren H; Chen N, Zou C. Abstract: SummaryWe propose a procedure based on a high-breakdown mean function estimator to detect outliers in functional data. The robust estimator is obtained from a clean subset of observations, excluding potential outliers, by minimizing the least-trimmed-squares projection coefficients after functional principal component analysis. A threshold rule based on the asymptotic distribution of the functional score-based distance robustly controls the false positive rate and detects outliers effectively. Further improvement in power can be achieved by adding a one-step reweighting procedure. The finite-sample performance of our method demonstrates satisfactory false positive and false negative rates compared with existing outlier detection methods for functional data. PubDate: 2017-03-27

Authors:Constantinou PP; Kokoszka PP, Reimherr MM. Abstract: SummarySeparability is a common simplifying assumption on the covariance structure of spatiotemporal functional data. We present three tests of separability, one a functional extension of the Monte Carlo likelihood method of Mitchell et al. (2006) and two based on quadratic forms. Our tests are based on asymptotic distributions of maximum likelihood estimators and do not require Monte Carlo simulation. The main theoretical contribution of this paper is the specification of the joint asymptotic distribution of these estimators, which can be used to derive many other tests. The main methodological finding is that one of the quadratic form methods, which we call a norm approach, emerges as a clear winner in terms of finite-sample performance in nearly every setting we considered. This approach focuses directly on the Frobenius distance between the spatiotemporal covariance function and its separable approximation. We demonstrate the efficacy of our methods via simulations and application to Irish wind data. PubDate: 2017-03-20

Authors:Binkiewicz NN; Vogelstein JT, Rohe KK. Abstract: SummaryBiological and social systems consist of myriad interacting units. The interactions can be represented in the form of a graph or network. Measurements of these graphs can reveal the underlying structure of these interactions, which provides insight into the systems that generated the graphs. Moreover, in applications such as connectomics, social networks, and genomics, graph data are accompanied by contextualizing measures on each node. We utilize these node covariates to help uncover latent communities in a graph, using a modification of spectral clustering. Statistical guarantees are provided under a joint mixture model that we call the node-contextualized stochastic blockmodel, including a bound on the misclustering rate. The bound is used to derive conditions for achieving perfect clustering. For most simulated cases, covariate-assisted spectral clustering yields results superior both to regularized spectral clustering without node covariates and to an adaptation of canonical correlation analysis. We apply our clustering method to large brain graphs derived from diffusion MRI data, using the node locations or neurological region membership as covariates. In both cases, covariate-assisted spectral clustering yields clusters that are easier to interpret neurologically. PubDate: 2017-03-19

Authors:Holmes CC; Walker SG. Abstract: SummaryBayesian robustness under model misspecification is a current area of active research. Among recent ideas is that of raising the likelihood function to a power. In this paper we discuss the choice of appropriate power and provide examples. PubDate: 2017-03-19

Authors:Chen S; Haziza D. Abstract: SummaryItem nonresponse in surveys is often treated through some form of imputation. We introduce multiply robust imputation in finite population sampling. This is closely related to multiple robustness, which extends double robustness. In practice, multiple nonresponse models and multiple imputation models may be fitted, each involving different subsets of covariates and possibly different link functions. An imputation procedure is said to be multiply robust if the resulting estimator is consistent when all models but one are misspecified. A jackknife variance estimator is proposed and shown to be consistent. Random and fractional imputation procedures are discussed. A simulation study suggests that the proposed estimation procedures have low bias and high efficiency. PubDate: 2017-03-17

Authors:Kosmidis II; Guolo AA, Varin CC. Abstract: SummaryRandom-effects models are frequently used to synthesize information from different studies in meta-analysis. While likelihood-based inference is attractive both in terms of limiting properties and of implementation, its application in random-effects meta-analysis may result in misleading conclusions, especially when the number of studies is small to moderate. The current paper shows how methodology that reduces the asymptotic bias of the maximum likelihood estimator of the variance component can also substantially improve inference about the mean effect size. The results are derived for the more general framework of random-effects meta-regression, which allows the mean effect size to vary with study-specific covariates. PubDate: 2017-03-08

Authors:Xiao Q Xu H. Abstract: SummaryMaximin distance Latin hypercube designs are widely used in computer experiments, yet their construction is challenging. Based on number theory and finite fields, we propose three algebraic methods to construct maximin distance Latin squares as special Latin hypercube designs. We develop lower bounds on their minimum distances. The resulting Latin squares and related Latin hypercube designs have larger minimum distances than existing ones, and are especially appealing for high-dimensional applications. PubDate: 2017-02-28

Authors:Ehm WW; Ovcharov EY. Abstract: SummaryDecompositions of the score of a forecast represent useful tools for assessing its performance. We consider local score decompositions permitting detailed forecast assessments across a spectrum of conditions of interest. We derive corrections to the bias of the decomposition components in the framework of point forecasts of quantile-type functionals, and illustrate their performance by simulation. Related bias corrections have thus far only been known for squared error criteria. PubDate: 2017-02-27