Papers

We construct exact confidence intervals for the average treatment effect in randomized experiments with binary outcomes using sequences of randomization tests. Our approach does not rely on large-sample approximations and is valid for all sample sizes. Under a balanced Bernoulli design or a matched-pairs design, we show that exact confidence intervals can be computed using only $O(\log n)$ randomization tests, yielding an exponential reduction in the number of tests compared to brute-force. We further prove an information-theoretic lower bound showing that this rate is optimal. In contrast, under balanced complete randomization, the most efficient known procedures require $O(n\log n)$ randomization tests (Aronow et al., 2023), establishing a sharp separation between these designs. In addition, we extend our algorithm to general Bernoulli designs using $O(n^2)$ randomization tests.

We study online alignment of large language models under misspecified preference feedback, where the observed preference oracle deviates from an ideal but unknown ground-truth oracle. The online LLM alignment problem is a bi-level reinforcement problem due to the coupling between data collection and policy updates. Recently, the problem has been reduced to tractable single-level objective in the SAIL (Self-Improving Efficient Online Alignment) framework. In this paper, we introduce a pointwise oracle uncertainty set in this problem and formulate an oracle-robust online alignment objective as a worst-case optimization problem. For log-linear policies, we show that this robust objective admits an exact closed-form decomposition into the original loss function plus an explicit sensitivity penalty. We develop projected stochastic composite updates for the resulting weakly convex objective and prove $\widetilde{O}(\varepsilon^{-2})$ oracle complexity for reaching approximate stationarity.

For large model spaces in linear regression with spike-and-slab priors, the potential entrapment of Markov chain Monte Carlo (MCMC)-based methods poses significant challenges in posterior computation. Existing maximum a posteriori (MAP)-based methods provide more computationally viable alternatives, but fail to perform tail heaviness estimation and uncertainty quantification. To address these problems, we propose a method that blends MAP estimation with MCMC-based stochastic search algorithms within an error framework comprising a combination of the hyperbolic and Student-t distributions. The hyperbolic distribution has the light-tailed normal and heavy-tailed Laplace distributions as limiting cases, but is thinner-tailed than the Student-t family. Including the Student-t distribution in the error density enables better adaptation to heavier tails. Amalgamating the two error densities thus ensures a model with more flexible tail behavior when faced with unknown tail thickness in the data, compared to MAP estimators with fixed levels of tail heaviness that assume the errors have a normal or Laplace distribution. Under this proposed error model, the current work develops a two-step expectation conditional maximization (ECM)-guided MCMC algorithm. First, we conduct an ECM-based posterior maximization to guide variable selection. We then execute a Gibbs sampler on the resulting ECM-guided model space for tail heaviness estimation and uncertainty quantification. Through simulation studies and benchmark real datasets, our proposed method is shown to exhibit several advantages in variable selection and uncertainty quantification over state-of-the-art MAP-based methods. To implement our proposed method, we developed the R package FlexBayesReg, available at https://github.com/shamriddha1998/FlexBayesReg.

Independent sample generation is the prevailing paradigm in modern diffusion-based generative models of AI. We ask a different question: can samples coordinate through shared population statistics to transport probability mass more efficiently? We introduce Mean-Field Path-Integral Diffusion (MF-PID), a framework in which samples are promoted to interacting agents whose drift depends self-consistently on the evolving population density. The coupling converts distribution matching into a McKean–Vlasov extension of the stochastic optimal transport problem, unifying generative modeling and multi-agent control under the same Hamilton–Jacobi–Bellman/Kolmogorov–Fokker–Planck duality. We identify two analytically tractable regimes: a Linear–Quadratic–Gaussian (LQG) benchmark in which the infinite-dimensional mean-field system reduces to a finite set of Riccati and linear ODEs, and a Gaussian-mixture regime governed by a piecewise-constant protocol that preserves closed-form solvability. For a quadratic interaction potential with schedule $\beta_t$ and zero base drift we prove that the self-consistent MF guidance is the exact linear interpolant between initial and target global means – a result that holds for arbitrary initial and target densities and any $\beta_t$. Applied to demand-response control of energy systems, where agents aggregated into an ensemble are energy consumers (e.g.\ thermal zones within a building), MF-PID achieves 19–24% reductions in cumulative control energy over independent-agent baselines while matching the prescribed terminal distribution exactly, and reveals how coordination redistributes actuation effort across heterogeneous sub-populations.

We study online Wasserstein distributionally robust optimization (DRO), where the ambiguity set is a Wasserstein ball centered at the empirical distribution and evolves with each observation. We formulate each round as a zero-sum game and develop an online distributional best-response algorithm that requires only one projected subgradient step per iteration. Combining regret analysis with measure concentration, we establish explicit, instance-dependent, and high-probability convergence rates to the optimal offline DRO value.

Explainable artificial intelligence promises to yield insights into relevant features, thereby enabling humans to examine and scrutinize machine learning models or even facilitating scientific discovery. Considering the widespread technique of Shapley values, we find that purely data-driven operationalization of multivariate feature importance is unsuitable for such purposes. Even for simple problems with two features, spurious associations due to collider bias and suppression arise from considering one feature only in the observational context of the other, which can lead to misinterpretations. Causal knowledge about the data-generating process is required to identify and correct such misleading feature attributions. We propose cc-Shapley (causal context Shapley), an interventional modification of conventional observational Shapley values leveraging knowledge of the data's causal structure, thereby analyzing the relevance of a feature in the causal context of the remaining features. We show theoretically that this eradicates spurious association induced by collider bias. We compare the behavior of Shapley and cc-Shapley values on various, synthetic, and real-world datasets. We observe nullification or reversal of associations compared to univariate feature importance when moving from observational to cc-Shapley.

Autoregressive models enable tractable sampling from learned probability distributions, but their performance critically depends on the variable ordering used in the factorization via complexities of the resulting conditional distributions. We propose to learn the Markov random field describing the underlying data, and use the inferred graphical model structure to construct optimized variable orderings. We illustrate our approach on two-dimensional image-like models where a structure-aware ordering leads to restricted conditioning sets, thereby reducing model complexity. Numerical experiments on Ising models with discrete data demonstrate that graph-informed orderings yield higher-fidelity generated samples compared to naive variable orderings.

Heterogeneous treatment effects (HTEs) are increasingly estimated using machine learning models that produce highly personalized predictions of treatment effects. In practice, however, predicted treatment effects are rarely interpreted, reported, or audited at the individual level but, instead, are often aggregated to broader subgroups, such as demographic segments, risk strata, or markets. We show that such aggregation can induce systematic bias of the group-level causal effect: even when models for predicting the individual-level conditional average treatment effect (CATE) are correctly specified and trained on data from randomized experiments, aggregating the predicted CATEs up to the group level does not, in general, recover the corresponding group average treatment effect (GATE). We develop a unified statistical framework to detect and mitigate this form of group bias in randomized experiments. We first define group bias as the discrepancy between the model-implied and experimentally identified GATEs, derive an asymptotically normal estimator, and then provide a simple-to-implement statistical test. For mitigation, we propose a shrinkage-based bias-correction, and show that the theoretically optimal and empirically feasible solutions have closed-form expressions. The framework is fully general, imposes minimal assumptions, and only requires computing sample moments. We analyze the economic implications of mitigating detected group bias for profit-maximizing personalized targeting, thereby characterizing when bias correction alters targeting decisions and profits, and the trade-offs involved. Applications to large-scale experimental data at major digital platforms validate our theoretical results and demonstrate empirical performance.

The validity of two-step or plug-in inference methods is questioned in the Bayesian framework. We study semi-parametric models where the plug-in of a non-parametrically modelled nuisance component is used. We show that when the nuisance and targeted parameters satisfy a Neyman orthogonal score property, the approach of cutting feedback through a two-step procedure is a valid way of conducting Bayesian inference. Our method relies on a non-parametric Bayesian formulation based on the Dirichlet process and the Bayesian bootstrap. We show that the marginal posterior of the targeted parameter exhibits good frequentist properties despite not accounting for the inferential uncertainty of the nuisance parameter. We adopt this approach in Bayesian causal inference problems where the nuisance propensity score model is estimated to obtain marginal inference for the treatment effect parameter, and demonstrate that a plug-in of the propensity score has a negligible effect on marginal posterior inference for the causal contrast. We investigate the absence of Neyman orthogonality and exploit our findings to show that in conventional two-step procedures, the posterior distribution converges under weaker restrictions than those needed in the frequentist sequel. For a simple family of useful scores, we demonstrate that even in the absence of Neyman orthogonality, the posterior distribution is asymptotically unchanged by the estimation of the nuisance parameter, merely provided the latter estimator is consistent.

We study gap-dependent regret for reinforcement learning with linear function approximation. While prior works have established gap-dependent guarantees in this setting, existing analyses do not apply to algorithms that achieve the nearly minimax-optimal worst-case regret bound $\tilde{O}(d\sqrt{H^3K})$, where $d$ is the feature dimension, $H$ is the horizon length, and $K$ is the number of episodes. We bridge this gap by establishing the first gap-dependent regret bound for the nearly minimax-optimal algorithm LSVI-UCB++ (He et al., 2023), with an expected regret bound $\tilde{O}(d^2H^3/\Delta_{\min}+d^6H^5)$, improving the dependence on both $d$ and $H$ in the leading gap-dependent term compared with previous results. To understand the exploration cost induced by optimism, we establish a structural lower bound $\Omega(d^2H^3/\Delta_{\min})$ for a broad class of algorithms based on persistent ellipsoidal optimism. When specialized to LSVI-UCB++, this result shows that the leading dependence of our upper bound on $d$, $H$, and $\Delta_{\min}$ is tight up to logarithmic factors. Beyond this algorithmic class, we establish a general gap-dependent lower bound $\Omega(dH^3\log K/\Delta_{\min})$ for arbitrary learning algorithms, showing that the logarithmic dependence on $K$ and the cubic dependence on $H$ are intrinsic to gap-dependent expected regret in linear MDPs. Together, our results substantially narrow the gap between upper and lower bounds and provide a sharper characterization of gap-dependent learning and optimism-based exploration with linear function approximation.

We study a discrete denoising diffusion framework that integrates a sample-efficient estimator of single-site conditionals with round-robin noising and denoising dynamics for generative modeling over discrete state spaces. Rather than approximating a discrete analog of a score function, our formulation treats single-site conditional probabilities as the fundamental objects that parameterize the reverse diffusion process. We employ a sample-efficient method known as Neural Interaction Screening Estimator (NeurISE) to estimate these conditionals in the diffusion dynamics. Controlled experiments on synthetic Ising models, MNIST, and scientific data sets produced by a D-Wave quantum annealer, synthetic Potts model and one dimensional quantum systems demonstrate the proposed approach. On the binary data sets, these experiments demonstrate that the proposed approach outperforms popular existing methods including ratio-based approaches, achieving improved performance in total variation, cross-correlations, and kernel density estimation metrics.

Accurate delivery delay prediction is critical for maintaining operational efficiency and customer satisfaction across modern supply chains. Yet the increasing complexity of logistics networks, spanning multimodal transportation, cross-country routing, and pronounced regional variability, makes this prediction task inherently challenging. This paper introduces a multi-task deep learning model for delivery delay duration prediction in the presence of significant imbalanced data, where delayed shipments are rare but operationally consequential. The model embeds high-dimensional shipment features with dedicated embedding layers for tabular data, and then uses a classification-then-regression strategy to predict the delivery delay duration for on-time and delayed shipments. Unlike sequential pipelines, this approach enables end-to-end training, improves the detection of delayed cases, and supports probabilistic forecasting for uncertainty-aware decision making. The proposed approach is evaluated on a large-scale real-world dataset from an industrial partner, comprising more than 10 million historical shipment records across four major source locations with distinct regional characteristics. The proposed model is compared with traditional machine learning methods. Experimental results show that the proposed method achieves a mean absolute error of 0.67-0.91 days for delayed-shipment predictions, outperforming single-step tree-based regression baselines by 41-64% and two-step classify-then-regress tree-based models by 15-35%. These gains demonstrate the effectiveness of the proposed model in operational delivery delay forecasting under highly imbalanced and heterogeneous conditions.

We construct a new class of infinite-dimensional diffusions with values in a generalized Kingman simplex with finitely many marks. The model describes the temporal evolution of the relative frequencies of infinitely many types that are labeled by a finite number $H$ of marks, but unlabeled within each mark. We first establish a blockwise skew-product representation for a finite-type Wright-Fisher diffusion, extending the aggregation-renormalization self-similarity property of Dirichlet laws. The decomposition separates an $H$-dimensional Wright-Fisher diffusion governing the evolving random mark masses, from $H$ Wright-Fisher diffusions, each run on its own random clock, which describe the evolution of the relative frequencies within each mark. After ranking the within-mark frequencies in decreasing order, we identify the distributional limit as the number of types per mark tends to infinity and we derive an explicit form of its infinitesimal generator on a suitable domain. The limiting diffusion admits the multiple Poisson-Dirichlet distribution as a stationary distribution; it recovers the infinitely-many-neutral-alleles diffusion when all types share the same mark and yields a diffusion on the Thoma simplex when there are two marks.

We study post-calibration uncertainty for trained ensembles of classifiers. Specifically, we consider both aleatoric (label noise) and epistemic (model) uncertainty. Among the most popular and widely used calibration methods in classification are temperature scaling (i.e., pool-then-calibrate) and conformal methods. However, the main shortcoming of these calibration methods is that they do not balance the proportion of aleatoric and epistemic uncertainty. Not balancing these uncertainties can severely misrepresent predictive uncertainty, leading to overconfident predictions in some input regions while being underconfident in others. To address this shortcoming, we present a simple but powerful calibration algorithm Joint Uncertainty Calibration (JUCAL) that jointly calibrates aleatoric and epistemic uncertainty. JUCAL jointly calibrates two constants to weight and scale epistemic and aleatoric uncertainties by optimizing the negative log-likelihood (NLL) on the validation/calibration dataset. JUCAL can be applied to any trained ensemble of classifiers (e.g., transformers, CNNs, or tree-based methods), with minimal computational overhead, without requiring access to the models' internal parameters. We experimentally evaluate JUCAL on various text classification tasks, for ensembles of varying sizes and with different ensembling strategies. Our experiments show that JUCAL significantly outperforms SOTA calibration methods across all considered classification tasks, reducing NLL and predictive set size by up to 15% and 20%, respectively. Interestingly, even applying JUCAL to an ensemble of size 5 can outperform temperature-scaled ensembles of size up to 50 in terms of NLL and predictive set size, resulting in up to 10 times smaller inference costs. Thus, we propose JUCAL as a new go-to method for calibrating ensembles in classification.

Inspired by behavioral science, we propose Behavior Learning (BL), a novel general-purpose machine learning framework that learns interpretable and identifiable optimization structures from data, ranging from single optimization problems to hierarchical compositions. It unifies predictive performance, intrinsic interpretability, and identifiability, with broad applicability to scientific domains involving optimization. BL parameterizes a compositional utility function built from intrinsically interpretable modular blocks, which induces a data distribution for prediction and generation. Each block represents and can be written in symbolic form as a utility maximization problem (UMP), a foundational paradigm in behavioral science and a universal framework of optimization. BL supports architectures ranging from a single UMP to hierarchical compositions, the latter modeling hierarchical optimization structures. Its smooth and monotone variant (IBL) guarantees identifiability. Theoretically, we establish the universal approximation property of BL, and analyze the M-estimation properties of IBL. Empirically, BL demonstrates strong predictive performance, intrinsic interpretability and scalability to high-dimensional data. Code: https://github.com/MoonYLiang/Behavior-Learning ; install via pip install blnetwork.

Conformal risk control is an extension of conformal prediction for controlling risk functions beyond miscoverage. The original algorithm controls the expected value of a loss that is monotonic in a one-dimensional parameter. Here, we present risk control guarantees for generic algorithms applied to possibly non-monotonic losses with multidimensional parameters. The guarantees depend on the stability of the algorithm – unstable algorithms have looser guarantees. We give applications of this technique to selective image classification, FDR and IOU control of tumor segmentations, and multigroup debiasing of recidivism predictions across overlapping race and sex groups using empirical risk minimization.

Diffusion language models (DLMs) have recently emerged as a promising alternative to autoregressive (AR) approaches, enabling parallel token generation beyond a rigid left-to-right order. Despite growing empirical success, the theoretical understanding of how unmasking schedules – which specify the order and size of unmasked tokens during sampling – affect generation quality remains limited. In this work, we introduce a distribution-agnostic unmasking schedule for DLMs that adapts to the (unknown) dependence structure of the target data distribution, without requiring any prior knowledge or hyperparameter tuning. In contrast to prior deterministic procedures that fix unmasking sizes, our method randomizes the number of tokens revealed at each iteration. We show that, for two specific parameter choices, the sampling convergence guarantees – measured by Kullback-Leibler (KL) divergence – scale as $\widetilde O(\mathsf{TC}/K)$ and $\widetilde O(\mathsf{DTC}/K)$ respectively. Here, $K$ is the number of iterations, and $\mathsf{TC}$ and $\mathsf{DTC}$ are the total correlation and dual total correlation of the target distribution, capturing the intrinsic dependence structure underlying the data. Importantly, our guarantees hold in the practically relevant parallel-sampling regime $K<L$ where $L$ is the token sequence length. These results significantly improve upon prior convergence theories and yield substantial sampling acceleration for low-complexity distributions. Overall, our findings unveil the adaptivity of DLMs to intrinsic data structures and shed light on the benefit of randomized unmasking sizes in inference schedule design.

Bonferroni's correction is a popular tool to address multiplicity but is notorious for its low power when tests are dependent. This paper proposes a practical modification of Bonferroni's correction when test statistics are jointly normal and exchangeable. This method is intuitive to practitioners and achieves higher power in sparse alternatives, as our simulations suggest. We also prove that this method successfully controls the family-wise error rate at any significance level.

We study compound decision theory from a nonparametric Bayesian perspective, with particular emphasis on their relationship to empirical Bayes (EB) procedures. Motivated by the sharp risk guarantees available for EB procedures based on the nonparametric maximum likelihood estimator (NPMLE), we investigate whether analogous guarantees can be established for fully Bayesian decision rules. In a class of Gaussian compound decision problems, we show that the fully Bayesian posterior mean achieves near-optimal risk. Moreover, it is admissible as a genuine Bayes rule, whereas the corresponding NPMLE plug-in rule is inadmissible. Simulations illustrate the performance of nonparametric Bayes procedures relative to common alternatives. As an application, we apply our methodology to Census tract-level estimates of economic mobility from the Opportunity Atlas.

To measure the degree of agreement between two observers that independently classify $n$ subjects within $K$ categories, it is common to use different kappa type coefficients, the most common of which is the $\kappa_C$ coefficient (Cohen's kappa). As $\kappa_C$ has some weaknesses -such as its poor performance with highly unbalanced marginal distributions-, the $\Delta$ coefficient is sometimes used, based on the $delta$ response model. This model allows us to obtain other parameters like: (a) the $\alpha_i$ contribution of each $i$ category to the value of the global agreement $\Delta=\sum \alpha_i$; and (b) the consistency $\mathcal{S}_i$ in the category $i$ (degree of agreement in the category $i$), a more appropriate parameter than the kappa value obtained by collapsing the data into the category $i$. It has recently been shown that the classic estimator $\hat{\kappa}_C$ underestimates $\kappa_C$, having obtained a new estimator $\hat{\kappa}_{CU}$ which is less biased. This article demonstrates that something similar happens to the known estimators $\hat{\Delta}$, $\hat{\alpha}_i$, and $\hat{\mathcal{S}}_i$ of $\Delta$, $\alpha_i$ and $\mathcal{S}_i$ (respectively), proposes new and less biased estimators $\hat{\Delta}_U$, $\hat{\alpha}_{iU}$, and $\hat{\mathcal{S}}_{iU}$, determines their variances, analyses the behaviour of all estimators, and concludes that the new estimators should be used when $n$ or $K$ are small (at least when $n\leq 50$ or $K\leq 3$). Additionally, the case where one of the raters is a gold standard is contemplated, in which situation two new parameters arise: the $conformity$ (the rater's capability to recognize a subject in the category $i$) and the $predictivity$ (the reliability of a response $i$ by the rater).

Pretraining and fine-tuning are central stages in modern machine learning systems. In practice, feature learning plays an important role across both stages: deep neural networks learn a broad range of useful features during pretraining and further refine those features during fine-tuning. However, an end-to-end theoretical understanding of how choices of initialization impact the ability to reuse and refine features during fine-tuning has remained elusive. Here we develop an analytical theory of the pretraining fine-tuning pipeline in diagonal linear networks, deriving exact expressions for the generalization error as a function of initialization parameters and task statistics. We find that different initialization choices place the network into four distinct fine-tuning regimes that are distinguished by their ability to support feature learning and reuse and therefore by the task statistics for which they are beneficial. In particular, a smaller initialization scale in earlier layers enables the network to both reuse and refine its features, leading to superior generalization on fine-tuning tasks that rely on a subset of pretraining features. We demonstrate empirically that the same initialization parameters impact generalization in ResNets trained on CIFAR-100 and SVHN as well as Transformers trained on modular arithmetic tasks. Overall, our results demonstrate an alytically how data and network initialization interact to shape fine-tuning generalization, highlighting an important role for the relative scale of initialization across different layers in enabling continued feature learning during fine-tuning.

P-splines provide a flexible and computationally efficient smoothing framework and are commonly used for derivative estimation in functional data. Including an additive penalty term in P-splines has been shown to improve estimates of derivatives. We propose a method which incorporates the fast covariance estimation (FACE) algorithm with an additive penalty in P-splines. The proposed method is used to estimate derivatives of covariance for functional data, which play an important role in derivative-based functional principal component analysis (FPCA). Following this, we provide an algorithm for estimating the eigenfunctions and their corresponding scores in derivative-based FPCA. For comparison, we evaluate our algorithm against an existing function FPCAder() in simulation. In addition, we extend the algorithm to multivariate cases, referred to as derivative multivariate functional principal component analysis (DMFPCA). DMFPCA is applied to joint angles in human movement data, where the derivative-based scores demonstrate strong performance in distinguishing locomotion tasks.

I–MR charts commonly estimate the process standard deviation $\sigma$ via the span-2 average moving range divided by the unbiasing constant $d_2$; unlike the unbiased sample standard deviation ($S/c_4$), this estimator depends on ordering through adjacency, so permuting a fixed sample changes it. We formalize this by introducing an independent uniformly random permutation and applying the law of total variance, yielding an exact decomposition into a values component (variance of the permutation mean) and an adjacency component (expected conditional variance over permutations). The permutation mean is order-invariant and equals $\GMD/d_2$, where $\GMD$ is the sample Gini mean difference. Under i.i.d.\ Normal sampling, both components admit closed forms; the adjacency fraction converges to $0.3813$, and the familiar asymptotic efficiency loss relative to $S/c_4$ is almost entirely an adjacency effect.

This paper develops a novel change point identification method for high-dimensional data using random projections. By projecting high-dimensional time series into a one-dimensional space, we are able to leverage the rich literature for univariate time series. We propose applying random projections multiple times and then combining the univariate test results using existing multiple comparison methods. Simulation results suggest that the proposed method tends to have better size and power, with more accurate location estimation. At the same time, random projections may introduce variability in the estimated locations. To enhance stability in practice, we recommend repeating the procedure, and using the mode of the estimated locations as a guide for the final change point estimate. An application to an Australian temperature dataset is presented. This study, though limited to the single change point setting, demonstrates the usefulness of random projections in change point analysis.

Uncertainty quantification is central to safe and efficient deployments of deep learning models, yet many computationally practical methods lack lacking rigorous theoretical motivation. Random network distillation (RND) is a lightweight technique that measures novelty via prediction errors against a fixed random target. While empirically effective, it has remained unclear what uncertainties RND measures and how its estimates relate to other approaches, e.g. Bayesian inference or deep ensembles. This paper establishes these missing theoretical connections by analyzing RND within the neural tangent kernel framework in the limit of infinite network width. Our analysis reveals two central findings in this limit: (1) The uncertainty signal from RND – its squared self-predictive error – is equivalent to the predictive variance of a deep ensemble. (2) By constructing a specific RND target function, we show that the RND error distribution can be made to mirror the centered posterior predictive distribution of Bayesian inference with wide neural networks. Based on this equivalence, we moreover devise a posterior sampling algorithm that generates i.i.d. samples from an exact Bayesian posterior predictive distribution using this modified Bayesian RND model. Collectively, our findings provide a unified theoretical perspective that places RND within the principled frameworks of deep ensembles and Bayesian inference, and offer new avenues for efficient yet theoretically grounded uncertainty quantification methods.

Accurate estimates of wind speeds at wind turbine hub heights are crucial for both wind resource assessment and day-to-day management of electricity grids with high renewable penetration. In the absence of direct measurements, parametric models are commonly used to extrapolate wind speeds from observed heights to turbine heights. Recent literature has proposed extensions to allow for spatially or temporally varying vertical wind gradients, that is, the rate at which wind speed changes with height. However, these approaches typically assume that reference height and hub height measurements are available at the same locations, which limits their applicability in operational settings where meteorological stations and wind farms are spatially separated. In this paper, we develop a two-step spatio-temporal framework to estimate turbine height wind speeds using only open-access observations from sparse meteorological stations. First, a non-parametric generalized additive model is trained on reanalysis data to perform vertical height extrapolation. Second, a spatial Gaussian process model interpolates these hub-height estimates to wind farm locations while explicitly propagating uncertainty from the height extrapolation stage. The proposed framework enables the construction of high-resolution, sub-hourly turbine-height wind speed time series and spatial wind maps using data available in real time, capabilities not provided by existing reanalysis products. We further provide calibrated uncertainty estimates that account for both vertical extrapolation and spatial interpolation errors. The approach is validated using hub-height measurements from seven operational wind farms in Ireland, demonstrating improved accuracy relative to ERA5 reanalysis while relying solely on real-time, open-access data.

Disruptions are an inherent feature of transportation systems, occurring unpredictably and with varying durations. Even after an incident is reported as resolved, disruptions can induce irregular train operations that generate substantial uncertainty in passenger waiting and travel times. Accurately forecasting post-disruption travel times therefore remains a critical challenge for transit operators and passenger information systems. This paper develops a Bayesian spatiotemporal modeling framework for post-disruption train travel times that explicitly captures train interactions, headway imbalance, and non-Gaussian distributional characteristics observed during recovery periods. The proposed model decomposes travel times into delay and journey components and incorporates a moving-average error structure to represent dependence between consecutive trains. Skew-normal and skew-$t$ distributions are employed to flexibly accommodate heteroskedasticity, skewness, and heavy-tailed behavior in post-disruption travel times. The framework is evaluated using high-resolution track-occupancy and disruption log data from the Montréal metro system, covering two lines in both travel directions. Empirical results indicate that post-disruption travel times exhibit pronounced distributional asymmetries that vary with traveled distance, as well as significant error dependence across trains. The proposed models consistently outperform baseline specifications in both point prediction accuracy and uncertainty quantification, with the skew-$t$ model demonstrating the most robust performance for longer journeys. These findings underscore the importance of incorporating both distributional flexibility and error dependence when forecasting post-disruption travel times in urban rail systems.

Modern data-driven applications increasingly rely on large, heterogeneous datasets collected across multiple sites. Differences in data availability, feature representation, and underlying populations often induce structured missingness, complicating efforts to transfer information from data-rich settings to those with limited data. Many transfer learning methods overlook this structure, limiting their ability to capture meaningful relationships across sites. We propose TransNEST (Transfer learning with Network Embeddings under STructured missingness), a framework that integrates graphical data from source and target sites with prior group structure to construct and refine network embeddings. TransNEST accommodates site-specific features, captures within-group heterogeneity and between-site differences adaptively, and improves embedding estimation under partial feature overlap. We establish the convergence rate for the TransNEST estimator and demonstrate strong finite-sample performance in simulations. We apply TransNEST to a multi-site electronic health record study, transferring feature embeddings from a general hospital system to a pediatric hospital system. Using a hierarchical ontology structure, TransNEST improves pediatric embeddings and supports more accurate pediatric knowledge extraction, achieving the best accuracy for identifying pediatric-specific relational feature pairs compared with benchmark methods.

Causal discovery problems use a set of observations to deduce causality between variables in the real world, typically to answer questions about biological or physical systems. These observations are often recorded at regular time intervals, determined by a user or a machine, depending on the experiment design. There is generally no guarantee that the timing of these recordings matches the timing of the underlying biological or physical events. In this paper, we examine the sensitivity of causal discovery methods to this potential mismatch. We consider empirical and theoretical evidence to understand how causal discovery performance is impacted by changes of sampling rate and window length. We demonstrate that both classical and recent causal discovery methods exhibit sensitivity to these hyperparameters, and we discuss how ideas from signal processing may help us understand these phenomena.

We present a family of generalized Hessian estimators of the objective using random direction stochastic approximation (RDSA) by utilizing only noisy function measurements. The form of each estimator and the order of the bias depend on the number of function measurements. In particular, we demonstrate that estimators with more function measurements exhibit lower-order estimation bias. We show the asymptotic unbiasedness of the estimators. We also perform asymptotic and non-asymptotic convergence analyses for stochastic Newton methods that incorporate our generalized Hessian estimators. Finally, we perform numerical experiments to validate our theoretical findings.