Search

Understanding directionality is crucial for identifying causal structures from observational data. A key challenge lies in detecting collider structures, where a $V$--structure is formed between a child node $Z$ receiving directed edges from parents $X$ and $Y$, denoted by $X \rightarrow Z \leftarrow Y$. Traditional causal discovery approaches, such as constraint-based and score-based structure learning algorithms, do not provide statistical inference on estimated pathways and are often sensitive to latent confounding. To overcome these issues, we introduce methodology to quantify directionality in collider structures using a pair of conditional asymmetry coefficients to simultaneously examine validity of the pathways $Y \rightarrow Z$ and $X \rightarrow Z$ in the collider structure. These coefficients are based on Shannon's differential entropy. Leveraging kernel-based conditional density estimation and a nonparametric smoothing technique, we utilise our proposed method to estimate collider structures and provide uncertainty quantification. Simulation studies demonstrate that our method outperforms existing structure learning algorithms in accurately identifying collider structures. We further apply our approach to investigate the role of blood pressure as a collider in epigenetic DNA methylation, uncovering novel insights into the genetic regulation of blood pressure. This framework represents a significant advancement in causal structure learning, offering a robust, nonparametric method for collider detection with practical applications in biostatistics and epidemiology.

Mediation analytics help examine if and how an intermediate variable mediates the influence of an exposure variable on an outcome of interest. Quantiles, rather than the mean, of an outcome are scientifically relevant to the comparison among specific subgroups in practical studies. Albeit some empirical studies available in the literature, there lacks a thorough theoretical investigation of quantile-based mediation analysis, which hinders practitioners from using such methods to answer important scientific questions. To address this significant technical gap, in this paper, we develop a quantile mediation analysis methodology to facilitate the identification, estimation, and testing of quantile mediation effects under a hypothesized directed acyclic graph. We establish two key estimands, quantile natural direct effect (qNDE) and quantile natural indirect effect (qNIE), in the counterfactual framework, both of which have closed-form expressions. To overcome the issue that the null hypothesis of no mediation effect is composite, we establish a powerful adaptive bootstrap method that is shown theoretically and numerically to achieve a proper type I error control. We illustrate the proposed quantile mediation analysis methodology through both extensive simulation experiments and a real-world dataset in that we investigate the mediation effect of lipidomic biomarkers for the influence of exposure to phthalates on early childhood obesity clinically diagnosed by 95\% percentile of body mass index.

Social network interference induces spillover effects from neighbors' exposures, and the complexity of statistical analysis increases when mediators are involved with network interference. To address various technical challenges, we develop a theoretical framework employing a structural graphical modeling approach to investigate both mediation and interference effects within network data. Our framework enables us to capture the multifaceted mechanistic pathways through which neighboring units' exposures and mediators exert direct and indirect influences on an individual's outcome. We extend the exposure mapping paradigm in the context of a random-effects network structural equation models (REN-SEM), establishing its capacity to delineate spillover effects of interest. Our proposed methodology contributions include maximum likelihood estimation for REN-SEM and inference procedures with theoretical guarantees. Such guarantees encompass consistent asymptotic variance estimators, derived under a non-i.i.d. asymptotic theory. The robustness and practical utility of our methodology are demonstrated through simulation experiments and a real-world data analysis of the Twitch Gamers Network Dataset, underscoring its effectiveness in capturing the intricate dynamics of network-mediated exposure effects. This work is the first to provide a rigorous theoretical framework and analytic toolboxes to the mediation analysis of network data, including a robust assessment on the interplay of mediation and interference.

Data sharing barriers are paramount challenges arising from multicenter clinical studies where multiple data sources are stored in a distributed fashion at different local study sites. Particularly in the case of time-to-event analysis when global risk sets are needed for the Cox proportional hazards model, access to a centralized database is typically necessary. Merging such data sources into a common data storage for a centralized statistical analysis requires a data use agreement, which is often time-consuming. Furthermore, the construction and distribution of risk sets to participating clinical centers for subsequent calculations may pose a risk of revealing individual-level information. We propose a new collaborative Cox model that eliminates the need for accessing the centralized database and constructing global risk sets but needs only the sharing of summary statistics with significantly smaller dimensions than risk sets. Thus, the proposed collaborative inference enjoys maximal protection of data privacy. We show theoretically and numerically that the new distributed proportional hazards model approach has little loss of statistical power when compared to the centralized method that requires merging the entire data. We present a renewable sieve method to establish large-sample properties for the proposed method. We illustrate its performance through simulation experiments and a real-world data example from patients with kidney transplantation in the Organ Procurement and Transplantation Network (OPTN) to understand the factors associated with the 5-year death-censored graft failure (DCGF) for patients who underwent kidney transplants in the US.

To evaluate a classification algorithm, it is common practice to plot the ROC curve using test data. However, the inherent randomness in the test data can undermine our confidence in the conclusions drawn from the ROC curve, necessitating uncertainty quantification. In this article, we propose an algorithm to construct confidence bands for the ROC curve, quantifying the uncertainty of classification on the test data in terms of sensitivity and specificity. The algorithm is based on a procedure called conformal prediction, which constructs individualized confidence intervals for the test set and the confidence bands for the ROC curve can be obtained by combining the individualized intervals together. Furthermore, we address both scenarios where the test data are either iid or non-iid relative to the observed data set and propose distinct algorithms for each case with valid coverage probability. The proposed method is validated through both theoretical results and numerical experiments.

We propose a unified class of generalized structural equation models (GSEMs) with data of mixed types in mediation analysis, including continuous, categorical, and count variables. Such models extend substantially the classical linear structural equation model to accommodate many data types arising from the application of mediation analysis. Invoking the hierarchical modeling approach, we specify GSEMs by a copula joint distribution of outcome variable, mediator and exposure variable, in which marginal distributions are built upon generalized linear models (GLMs) with confounding factors. We discuss the identifiability conditions for the causal mediation effects in the counterfactual paradigm as well as the issue of mediation leakage, and develop an asymptotically efficient profile maximum likelihood estimation and inference for two key mediation estimands, natural direct effect and natural indirect effect, in different scenarios of mixed data types. The proposed new methodology is illustrated by a motivating epidemiological study that aims to investigate whether the tempo of reaching infancy BMI peak (delay or on time), an important early life growth milestone, may mediate the association between prenatal exposure to phthalates and pubertal health outcomes., Comment: 33 pages, 3 figures, 3 tables

In many practical studies, learning directionality between a pair of variables is of great interest while notoriously hard, especially for mechanistic relationships. This paper presents a method that examines directionality in exposure-outcome pairs when a priori assumptions about their relative ordering are unavailable. We propose a coefficient of asymmetry to quantify directional asymmetry using Shannon's entropy and propose a statistical estimation and inference framework for said estimand. Large-sample theoretical guarantees are established through data-splitting and cross-fitting techniques. The proposed methodology is extended to allow both measured confounders and contamination in outcome measurements. The methodology is extensively evaluated through extensive simulation studies, a benchmark dataset, and a real data application., Comment: Code available: https://github.com/soumikp/2025_jmlr

Mediation analysis aims to assess if, and how, a certain exposure influences an outcome of interest through intermediate variables. This problem has recently gained a surge of attention due to the tremendous need for such analyses in scientific fields. Testing for the mediation effect is greatly challenged by the fact that the underlying null hypothesis (i.e. the absence of mediation effects) is composite. Most existing mediation tests are overly conservative and thus underpowered. To overcome this significant methodological hurdle, we develop an adaptive bootstrap testing framework that can accommodate different types of composite null hypotheses in the mediation pathway analysis. Applied to the product of coefficients (PoC) test and the joint significance (JS) test, our adaptive testing procedures provide type I error control under the composite null, resulting in much improved statistical power compared to existing tests. Both theoretical properties and numerical examples of the proposed methodology are discussed.

It is common practice to use data containing personal information to build predictive models in the framework of empirical risk minimization (ERM). While these models can be highly accurate in prediction, sharing the results from these models trained on sensitive data may be susceptible to privacy attacks. Differential privacy (DP) is an appealing framework for addressing such data privacy issues by providing mathematically provable bounds on the privacy loss incurred when releasing information from sensitive data. Previous work has primarily concentrated on applying DP to unweighted ERM. We consider weighted ERM (wERM), an important generalization, where each individual's contribution to the objective function can be assigned varying weights. We propose the first differentially private algorithm for general wERM, with theoretical DP guarantees. Extending the existing DP-ERM procedures to wERM creates a pathway for deriving privacy-preserving learning methods for individualized treatment rules, including the popular outcome weighted learning (OWL). We evaluate the performance of the DP-wERM framework applied to OWL in both simulation studies and in a real clinical trial. All empirical results demonstrate the feasibility of training OWL models via wERM with DP guarantees while maintaining sufficiently robust model performance, providing strong evidence for the practicality of implementing the proposed privacy-preserving OWL procedure in real-world scenarios involving sensitive data., Comment: 29 pages, 1 figure, and 1 table for the main manuscript; 10 pages, 4 figures, and 1 table for the supplementary materials

As a fundamental concept in information theory, mutual information ($MI$) has been commonly applied to quantify association between random vectors. Most existing nonparametric estimators of $MI$ have unstable statistical performance since they involve parameter tuning. We develop a consistent and powerful estimator, called fastMI, that does not incur any parameter tuning. Based on a copula formulation, fastMI estimates $MI$ by leveraging Fast Fourier transform-based estimation of the underlying density. Extensive simulation studies reveal that fastMI outperforms state-of-the-art estimators with improved estimation accuracy and reduced run time for large data sets. fastMI provides a powerful test for independence that exhibits satisfactory type I error control. Anticipating that it will be a powerful tool in estimating mutual information in a broad range of data, we develop an R package fastMI for broader dissemination., Comment: 5 tables, 5 figures

Causal investigations in observational studies pose a great challenge in research where randomized trials or intervention-based studies are not feasible. We develop an information geometric causal discovery and inference framework of "predictive asymmetry". For $(X, Y)$, predictive asymmetry enables assessment of whether $X$ is more likely to cause $Y$ or vice-versa. The asymmetry between cause and effect becomes particularly simple if $X$ and $Y$ are deterministically related. We propose a new metric called the Directed Mutual Information ($DMI$) and establish its key statistical properties. $DMI$ is not only able to detect complex non-linear association patterns in bivariate data, but also is able to detect and infer causal relations. Our proposed methodology relies on scalable non-parametric density estimation using Fourier transform. The resulting estimation method is manyfold faster than the classical bandwidth-based density estimation. We investigate key asymptotic properties of the $DMI$ methodology and a data-splitting technique is utilized to facilitate causal inference using the $DMI$. Through simulation studies and an application, we illustrate the performance of $DMI$.

Data sharing barriers are paramount challenges arising from multicenter clinical trials where multiple data sources are stored in a distributed fashion at different local study sites. Merging such data sources into a common data storage for a centralized statistical analysis requires a data use agreement, which is often time-consuming. Data merging may become more burdensome when causal inference is of primary interest because propensity score modeling involves combining many confounding variables, and systematic incorporation of this additional modeling in meta-analysis has not been thoroughly investigated in the literature. We propose a new causal inference framework that avoids the merging of subject-level raw data from multiple sites but needs only the sharing of summary statistics. The proposed collaborative inference enjoys maximal protection of data privacy and minimal sensitivity to unbalanced data distributions across data sources. We show theoretically and numerically that the new distributed causal inference approach has little loss of statistical power compared to the centralized method that requires merging the entire data. We present large-sample properties and algorithms for the proposed method. We illustrate its performance by simulation experiments and a real-world data example on a multicenter clinical trial of basal insulin treatment for reducing the risk of post-transplantation diabetes among kidney-transplant patients.

Fusing regression coefficients into homogenous groups can unveil those coefficients that share a common value within each group. Such groupwise homogeneity reduces the intrinsic dimension of the parameter space and unleashes sharper statistical accuracy. We propose and investigate a new combinatorial grouping approach called $L_0$-Fusion that is amenable to mixed integer optimization (MIO). On the statistical aspect, we identify a fundamental quantity called grouping sensitivity that underpins the difficulty of recovering the true groups. We show that $L_0$-Fusion achieves grouping consistency under the weakest possible requirement of the grouping sensitivity: if this requirement is violated, then the minimax risk of group misspecification will fail to converge to zero. Moreover, we show that in the high-dimensional regime, one can apply $L_0$-Fusion coupled with a sure screening set of features without any essential loss of statistical efficiency, while reducing the computational cost substantially. On the algorithmic aspect, we provide a MIO formulation for $L_0$-Fusion along with a warm start strategy. Simulation and real data analysis demonstrate that $L_0$-Fusion exhibits superiority over its competitors in terms of grouping accuracy.

Two dominant distributed computing strategies have emerged to overcome the computational bottleneck of supervised learning with big data: parallel data processing in the MapReduce paradigm and serial data processing in the online streaming paradigm. Despite the two strategies' common divide-and-combine approach, they differ in how they aggregate information, leading to different trade-offs between statistical and computational performance. In this paper, we propose a new hybrid paradigm, termed a Parallel-and-Stream Accelerator (PASA), that uses the strengths of both strategies for computationally fast and statistically efficient supervised learning. PASA's architecture nests online streaming processing into each distributed and parallelized data process in a MapReduce framework. PASA leverages the advantages and mitigates the disadvantages of both the MapReduce and online streaming approaches to deliver a more flexible paradigm satisfying practical computing needs. We study the analytic properties and computational complexity of PASA, and detail its implementation for two key statistical learning tasks. We illustrate its performance through simulations and a large-scale data example building a prediction model for online purchases from advertising data., Comment: 22 pages, 3 figures

It is of importance to develop statistical techniques to analyze high-dimensional data in the presence of both complex dependence and possible outliers in real-world applications such as imaging data analyses. We propose a new robust high-dimensional regression with coefficient thresholding, in which an efficient nonconvex estimation procedure is proposed through a thresholding function and the robust Huber loss. The proposed regularization method accounts for complex dependence structures in predictors and is robust against outliers in outcomes. Theoretically, we analyze rigorously the landscape of the population and empirical risk functions for the proposed method. The fine landscape enables us to establish both {statistical consistency and computational convergence} under the high-dimensional setting. The finite-sample properties of the proposed method are examined by extensive simulation studies. An illustration of real-world application concerns a scalar-on-image regression analysis for an association of psychiatric disorder measured by the general factor of psychopathology with features extracted from the task functional magnetic resonance imaging data in the Adolescent Brain Cognitive Development study., Comment: 38 pages

This paper develops an incremental learning algorithm based on quadratic inference function (QIF) to analyze streaming datasets with correlated outcomes such as longitudinal data and clustered data. We propose a renewable QIF (RenewQIF) method within a paradigm of renewable estimation and incremental inference, in which parameter estimates are recursively renewed with current data and summary statistics of historical data, but with no use of any historical subject-level raw data. We compare our renewable estimation method with both offline QIF and offline generalized estimating equations (GEE) approach that process the entire cumulative subject-level data, and show theoretically and numerically that our renewable procedure enjoys statistical and computational efficiency. We also propose an approach to diagnose the homogeneity assumption of regression coefficients via a sequential goodness-of-fit test as a screening procedure on occurrences of abnormal data batches. We implement the proposed methodology by expanding existing Spark's Lambda architecture for the operation of statistical inference and data quality diagnosis. We illustrate the proposed methodology by extensive simulation studies and an analysis of streaming car crash datasets from the National Automotive Sampling System-Crashworthiness Data System (NASS CDS).

Sparse representation has been widely used in data compression, signal and image denoising, dimensionality reduction and computer vision. While overcomplete dictionaries are required for sparse representation of multidimensional data, orthogonal bases represent one-dimensional data well. In this paper, we propose a data-driven sparse representation using orthonormal bases under the lossless compression constraint. We show that imposing such constraint under the Minimum Description Length (MDL) principle leads to a unique and optimal sparse representation for one-dimensional data, which results in discriminative features useful for data discovery.

We propose a distributed quadratic inference function framework to jointly estimate regression parameters from multiple potentially heterogeneous data sources with correlated vector outcomes. The primary goal of this joint integrative analysis is to estimate covariate effects on all outcomes through a marginal regression model in a statistically and computationally efficient way. We develop a data integration procedure for statistical estimation and inference of regression parameters that is implemented in a fully distributed and parallelized computational scheme. To overcome computational and modeling challenges arising from the high-dimensional likelihood of the correlated vector outcomes, we propose to analyze each data source using Qu, Lindsay and Li (2000)'s quadratic inference functions, and then to jointly reestimate parameters from each data source by accounting for correlation between data sources using a combined meta-estimator in a similar spirit to Hansen (1982)'s generalised method of moments. We show both theoretically and numerically that the proposed method yields efficiency improvements and is computationally fast. We illustrate the proposed methodology with the joint integrative analysis of the association between smoking and metabolites in a large multi-cohort study and provide an R package for ease of implementation., Comment: 26 pages, 4 figures, 4 tables

$\textbf{Objective}$: To develop a multi-channel device event segmentation and feature extraction algorithm that is robust to changes in data distribution. $\textbf{Methods}$: We introduce an adaptive transfer learning algorithm to classify and segment events from non-stationary multi-channel temporal data. Using a multivariate hidden Markov model (HMM) and Fisher's linear discriminant analysis (FLDA) the algorithm adaptively adjusts to shifts in distribution over time. The proposed algorithm is unsupervised and learns to label events without requiring $\textit{a priori}$ information about true event states. The procedure is illustrated on experimental data collected from a cohort in a human viral challenge (HVC) study, where certain subjects have disrupted wake and sleep patterns after exposure to a H1N1 influenza pathogen. $\textbf{Results}$: Simulations establish that the proposed adaptive algorithm significantly outperforms other event classification methods. When applied to early time points in the HVC data the algorithm extracts sleep/wake features that are predictive of both infection and infection onset time. $\textbf{Conclusion}$: The proposed transfer learning event segmentation method is robust to temporal shifts in data distribution and can be used to produce highly discriminative event-labeled features for health monitoring. $\textbf{Significance}$: Our integrated multisensor signal processing and transfer learning method is applicable to many ambulatory monitoring applications.

This paper presents a unified framework for supervised learning and inference procedures using the divide-and-conquer approach for high-dimensional correlated outcomes. We propose a general class of estimators that can be implemented in a fully distributed and parallelized computational scheme. Modelling, computational and theoretical challenges related to high-dimensional correlated outcomes are overcome by dividing data at both outcome and subject levels, estimating the parameter of interest from blocks of data using a broad class of supervised learning procedures, and combining block estimators in a closed-form meta-estimator asymptotically equivalent to estimates obtained by Hansen (1982)'s generalized method of moments (GMM) that does not require the entire data to be reloaded on a common server. We provide rigorous theoretical justifications for the use of distributed estimators with correlated outcomes by studying the asymptotic behaviour of the combined estimator with fixed and diverging number of data divisions. Simulations illustrate the finite sample performance of the proposed method, and we provide an R package for ease of implementation., Comment: 49 pages, 1 figure

While Internet of Things (IoT) devices and sensors create continuous streams of information, Big Data infrastructures are deemed to handle the influx of data in real-time. One type of such a continuous stream of information is time series data. Due to the richness of information in time series and inadequacy of summary statistics to encapsulate structures and patterns in such data, development of new approaches to learn time series is of interest. In this paper, we propose a novel method, called pattern tree, to learn patterns in the times-series using a binary-structured tree. While a pattern tree can be used for many purposes such as lossless compression, prediction and anomaly detection, in this paper we focus on its application in time series estimation and forecasting. In comparison to other methods, our proposed pattern tree method improves the mean squared error of estimation.

This paper is motivated by a regression analysis of electroencephalography (EEG) neuroimaging data with high-dimensional correlated responses with multi-level nested correlations. We develop a divide-and-conquer procedure implemented in a fully distributed and parallelized computational scheme for statistical estimation and inference of regression parameters. Despite significant efforts in the literature, the computational bottleneck associated with high-dimensional likelihoods prevents the scalability of existing methods. The proposed method addresses this challenge by dividing responses into subvectors to be analyzed separately and in parallel on a distributed platform using pairwise composite likelihood. Theoretical challenges related to combining results from dependent data are overcome in a statistically efficient way using a meta-estimator derived from Hansen's generalized method of moments. We provide a rigorous theoretical framework for efficient estimation, inference, and goodness-of-fit tests. We develop an R package for ease of implementation. We illustrate our method's performance with simulations and the analysis of the EEG data, and find that iron deficiency is significantly associated with two auditory recognition memory related potentials in the left parietal-occipital region of the brain., Comment: 35 pages, 5 figures, 3 tables

Simultaneous inference after model selection is of critical importance to address scientific hypotheses involving a set of parameters. In this paper, we consider high-dimensional linear regression model in which a regularization procedure such as LASSO is applied to yield a sparse model. To establish a simultaneous post-model selection inference, we propose a method of contraction and expansion (MOCE) along the line of debiasing estimation that enables us to balance the bias-and-variance trade-off so that the super-sparsity assumption may be relaxed. We establish key theoretical results for the proposed MOCE procedure from which the expanded model can be selected with theoretical guarantees and simultaneous confidence regions can be constructed by the joint asymptotic normal distribution. In comparison with existing methods, our proposed method exhibits stable and reliable coverage at a nominal significance level with substantially less computational burden, and thus it is trustworthy for its application in solving real-world problems.

Objective: To develop multisensor-wearable-device sleep monitoring algorithms that are robust to health disruptions affecting sleep patterns. Methods: We develop an unsupervised transfer learning algorithm based on a multivariate hidden Markov model and Fisher's linear discriminant analysis, adaptively adjusting to sleep pattern shift by training on dynamics of sleep/wake states. The proposed algorithm operates, without requiring a priori information about true sleep/wake states, by establishing an initial training set with hidden Markov model and leveraging a taper window mechanism to learn the sleep pattern in an incremental fashion. Our domain-adaptation algorithm is applied to a dataset collected in a human viral challenge study to identify sleep/wake periods of both uninfected and infected participants. Results: The algorithm successfully detects sleep/wake sessions in subjects whose sleep patterns are disrupted by respiratory infection (H3N2 flu virus). Pre-symptomatic features based on the detected periods are found to be strongly predictive of both infection status (AUC = 0.844) and infection onset time (AUC = 0.885), indicating the effectiveness and usefulness of the algorithm. Conclusion: Our method can effectively detect sleep/wake states in the presence of sleep pattern shift. Significance: Utilizing integrated multisensor signal processing and adaptive training schemes, our algorithm is able to capture key sleep patterns in ambulatory monitoring, leading to better automated sleep assessment and prediction.

The theory of statistical inference along with the strategy of divide-and-conquer for large- scale data analysis has recently attracted considerable interest due to great popularity of the MapReduce programming paradigm in the Apache Hadoop software framework. The central analytic task in the development of statistical inference in the MapReduce paradigm pertains to the method of combining results yielded from separately mapped data batches. One seminal solution based on the confidence distribution has recently been established in the setting of maximum likelihood estimation in the literature. This paper concerns a more general inferential methodology based on estimating functions, termed as the Rao-type confidence distribution, of which the maximum likelihood is a special case. This generalization provides a unified framework of statistical inference that allows regression analyses of massive data sets of important types in a parallel and scalable fashion via a distributed file system, including longitudinal data analysis, survival data analysis, and quantile regression, which cannot be handled using the maximum likelihood method. This paper investigates four important properties of the proposed method: computational scalability, statistical optimality, methodological generality, and operational robustness. In particular, the proposed method is shown to be closely connected to Hansen's generalized method of moments (GMM) and Crowder's optimality. An interesting theoretical finding is that the asymptotic efficiency of the proposed Rao-type confidence distribution estimator is always greater or equal to the estimator obtained by processing the full data once. All these properties of the proposed method are illustrated via numerical examples in both simulation studies and real-world data analyses.

We propose a distributed method for simultaneous inference for datasets with sample size much larger than the number of covariates, i.e., N >> p, in the generalized linear models framework. When such datasets are too big to be analyzed entirely by a single centralized computer, or when datasets are already stored in distributed database systems, the strategy of divide-and-combine has been the method of choice for scalability. Due to partition, the sub-dataset sample sizes may be uneven and some possibly close to p, which calls for regularization techniques to improve numerical stability. However, there is a lack of clear theoretical justification and practical guidelines to combine results obtained from separate regularized estimators, especially when the final objective is simultaneous inference for a group of regression parameters. In this paper, we develop a strategy to combine bias-corrected lasso-type estimates by using confidence distributions. We show that the resulting combined estimator achieves the same estimation efficiency as that of the maximum likelihood estimator using the centralized data. As demonstrated by simulated and real data examples, our divide-and-combine method yields nearly identical inference as the centralized benchmark., Comment: This paper is published in Journal of Multivariate Analysis