arXiv subjects

Xinkai Zhou

Publications and source records attributed to Xinkai Zhou.

9 recordsLinked to original sources

arXiv2026

FLAME: A Model for Duration-Dependent Risk Accumulation in Episodic Temporal Exposures

Emerging technologies enable continuous monitoring of temporal exposures to disease risk factors, leading to complex exposure processes characterized by subject-specific numbers and durations of exposure episodes. A key scientific question is how the number and duration of such episodes influence disease risk. Existing methods typically rely on scalar summaries or time-indexed representations and are not naturally suited to model duration-dependent risk accumulation at the episode level. We introduce the FLexible Accumulation ModEl (FLAME), a semiparametric model for risk accumulation at the level of individual exposure episodes, with duration as the primary driver of risk. FLAME is motivated by and applied to quantifying the association between the duration of intraoperative hypotension and acute kidney injury (AKI) following cardiac surgery. The estimated risk accumulation function reveals that, although 60 one-minute hypotensive episodes are associated with an AKI probability of 0.24, a single sustained 60-minute episode increases that probability to 0.33, representing a 38% increase despite identical total duration. These findings provide actionable insights for intraoperative hemodynamic management and demonstrate the importance of accounting for episodic exposure patterns. While motivated by cardiac surgery, FLAME is broadly applicable to other settings involving high-resolution temporal exposures. An R package, flameRisk, is provided to facilitate application of the method in practice.

stat.ME↗

arXiv2025

Function on Scalar Regression with Complex Survey Designs

Large health surveys increasingly collect high-dimensional functional data from wearable devices, and function on scalar regression (FoSR) is often used to quantify the relationship between these functional outcomes and scalar covariates such as age and sex. However, existing methods for FoSR fail to account for complex survey design. We introduce inferential methods for FoSR for studies with complex survey designs. The method combines fast univariate inference (FUI) developed for functional data outcomes and survey sampling inferential methods developed for scalar outcomes. Our approach consists of three steps: (1) fit survey weighted GLMs at each point along the functional domain, (2) smooth coefficients along the functional domain, and (3) use balanced repeated replication (BRR) or the Rao-Wu-Yue-Beaumont (RWYB) bootstrap to obtain pointwise and joint confidence bands for the functional coefficients. The method is motivated by association studies between continuous physical activity data and covariates collected in the National Health and Nutrition Examination Survey (NHANES). A first-of-its-kind analytical simulation study and empirical simulation using the NHANES data demonstrates that our method performs better than existing methods that do not account for the survey structure. Finally, application of the method in NHANES shows the practical implications of accounting for survey structure. The method is implemented in the R package svyfosr.

stat.ME↗

arXiv2025

An AI-Driven Multimodal Smart Home Platform for Continuous Monitoring and Assistance in Post-Stroke Motor Impairment

At-home rehabilitation for post-stroke patients presents significant challenges, as continuous, personalized care is often limited outside clinical settings. Moreover, the lack of integrated solutions capable of simultaneously monitoring motor recovery and providing intelligent assistance in home environments hampers rehabilitation outcomes. Here, we present a multimodal smart home platform designed for continuous, at-home rehabilitation of post-stroke patients, integrating wearable sensing, ambient monitoring, and adaptive automation. A plantar pressure insole equipped with a machine learning pipeline classifies users into motor recovery stages with up to 94\% accuracy, enabling quantitative tracking of walking patterns during daily activities. An optional head-mounted eye-tracking module, together with ambient sensors such as cameras and microphones, supports seamless hands-free control of household devices with a 100\% success rate and sub-second response time. These data streams are fused locally via a hierarchical Internet of Things (IoT) architecture, ensuring low latency and data privacy. An embedded large language model (LLM) agent, Auto-Care, continuously interprets multimodal data to provide real-time interventions -- issuing personalized reminders, adjusting environmental conditions, and notifying caregivers. Implemented in a post-stroke context, this integrated smart home platform increased mean user satisfaction from 3.9 $\pm$ 0.8 in conventional home environments to 8.4 $\pm$ 0.6 with the full system ($n=20$). Beyond stroke, the system offers a scalable, patient-centered framework with potential for long-term use in broader neurorehabilitation and aging-in-place applications.

cs.HC↗

arXiv2025

Fast Bayesian Functional Principal Components Analysis

Functional Principal Components Analysis (FPCA) is a widely used analytic tool for dimension reduction of functional data. Traditional implementations of FPCA estimate the principal components from the data, then treat these estimates as fixed in subsequent analyses. To account for the uncertainty of PC estimates, we propose FAST, a fully-Bayesian FPCA with three core components: (1) projection of eigenfunctions onto an orthonormal spline basis; (2) efficient sampling of the orthonormal spline coefficient matrix using a parameter expansion scheme based on polar decomposition; and (3) ordering eigenvalues during sampling. Extensive simulation studies show that FAST is very stable and performs better compared to existing methods. FAST is motivated by and applied to a study of the variability in mealtime glucose from the Dietary Approaches to Stop Hypertension for Diabetes Continuous Glucose Monitoring (DASH4D CGM) study. All relevant STAN code and simulation routines are available as supplementary material.

stat.ME↗

arXiv2025

Generalized Multilevel Functional Principal Component Analysis with Application to NHANES Active Inactive Patterns

Between 2011 and 2014 NHANES collected objectively measured physical activity data using wrist-worn accelerometers for tens of thousands of individuals for up to seven days. In this study, we analyze minute-level indicators of being active, which can be viewed as binary (since each minute is either active or inactive), multilevel (because there are multiple days of data for each participant), and functional data (because the within-day measurements can be viewed as a function of time). To identify both within- and between-participant directions of variation in these data, we introduce Generalized Multilevel Functional Principal Component Analysis (GM-FPCA), an approach based on the dimension reduction of the linear predictor. Our results indicate that specific activity patterns captured by GM-FPCA are strongly associated with mortality risk. Extensive simulation studies demonstrate that GM-FPCA accurately estimates model parameters, is computationally stable, and scales up with the number of study participants, visits, and observations per visit. R code for implementing the method is provided.

stat.ME↗

arXiv2025

Prediction Inference Using Generalized Functional Mixed Effects Models

We introduce inferential methods for prediction based on functional random effects in generalized functional mixed effects models. This is similar to the inference for random effects in generalized linear mixed effects models (GLMMs), but for functional instead of scalar outcomes. The method combines: (1) local GLMMs to extract initial estimators of the functional random components on the linear predictor scale; (2) structural functional principal components analysis (SFPCA) for dimension reduction; and (3) global Bayesian multilevel model conditional on the eigenfunctions for inference on the functional random effects. Extensive simulations demonstrate excellent coverage properties of credible intervals for the functional random effects in a variety of scenarios and for different data sizes. To our knowledge, this is the first time such simulations are conducted and reported, likely because prediction inference was not viewed as a priority and existing methods are too slow to calculate coverage. Methods are implemented in a reproducible R package and demonstrated using the NHANES 2011-2014 accelerometry data.

stat.ME↗

arXiv2024

Generalized Conditional Functional Principal Component Analysis

We propose generalized conditional functional principal components analysis (GC-FPCA) for the joint modeling of the fixed and random effects of non-Gaussian functional outcomes. The method scales up to very large functional data sets by estimating the principal components of the covariance matrix on the linear predictor scale conditional on the fixed effects. This is achieved by combining three modeling innovations: (1) fit local generalized linear mixed models (GLMMs) conditional on covariates in windows along the functional domain; (2) conduct a functional principal component analysis (FPCA) on the person-specific functional effects obtained by assembling the estimated random effects from the local GLMMs; and (3) fit a joint functional mixed effects model conditional on covariates and the estimated principal components from the previous step. GC-FPCA was motivated by modeling the minute-level active/inactive profiles over the day ($1{,}440$ 0/1 measurements per person) for $8{,}700$ study participants in the National Health and Nutrition Examination Survey (NHANES) 2011-2014. We show that state-of-the-art approaches cannot handle data of this size and complexity, while GC-FPCA can.

stat.ME↗

arXiv2023

Proximal MCMC for Bayesian Inference of Constrained and Regularized Estimation

This paper advocates proximal Markov Chain Monte Carlo (ProxMCMC) as a flexible and general Bayesian inference framework for constrained or regularized estimation. Originally introduced in the Bayesian imaging literature, ProxMCMC employs the Moreau-Yosida envelope for a smooth approximation of the total-variation regularization term, fixes variance and regularization strength parameters as constants, and uses the Langevin algorithm for the posterior sampling. We extend ProxMCMC to be fully Bayesian by providing data-adaptive estimation of all parameters including the regularization strength parameter. More powerful sampling algorithms such as Hamiltonian Monte Carlo are employed to scale ProxMCMC to high-dimensional problems. Analogous to the proximal algorithms in optimization, ProxMCMC offers a versatile and modularized procedure for conducting statistical inference on constrained and regularized problems. The power of ProxMCMC is illustrated on various statistical estimation and machine learning tasks, the inference of which is traditionally considered difficult from both frequentist and Bayesian perspectives.

stat.ME↗

arXiv2022

Star discrepancy for new stratified random sampling I: optimal expected star discrepancy

We introduce a class of convex equivolume partitions. Expected star discrepancy results are compared for stratified samples under these partitions, including simple random samples. There are four main parts of our results. First, among these newly designed partitions, there is one that minimizes the expected star discrepancy, thus we partly answer an open question in [F. Pausinger, S. Steinerberger, J. Complex. 2016]. Second, there are an infinite number of such class of partitions, which generate point sets with smaller expected discrepancy than classical jittered sampling for large sampling number, leading to an open question in [M. Kiderlen, F. Pausinger, Monatsh. Math. 2021] being solved. Third, we prove a strong partition principle and generalize the expected star discrepancy under these partition models from $L_2-$discrepancy to star discrepancy, hence an open question in [M. Kiderlen, F. Pausinger, J. Complex. 2021] is answered. In the end, optimal expected star discrepancy upper bound under this class of partitions is given, which is better than using jittered sampling.

math.ST↗