In real-life scenarios, a Reinforcement Learning (RL) agent aiming to maximise their reward, must often also behave in a safe manner, including at training time. Thus, much attention in recent years has been given to Safe RL, where an agent aims to learn an optimal policy among all policies that satisfy a given safety constraint. However, strict safety guarantees are often provided through approaches based on linear programming, and thus have limited scaling. In this paper we present a new, scalable method, which enjoys strict formal guarantees for Safe RL, in the case where the safety dynamics of the Markov Decision Process (MDP) are known, and safety is defined as an undiscounted probabilistic avoidance property. Our approach is based on state-augmentation of the MDP, and on the design of a shield that restricts the actions available to the agent. We show that our approach provides a strict formal safety guarantee that the agent stays safe at training and test time. Furthermore, we demonstrate that our approach is viable in practice through experimental evaluation.

Human physiological signals tend to exhibit both global and local structures: the former are shared across a population, while the latter reflect inter-individual variability. For instance, kinetic measurements of the gait cycle during locomotion present common characteristics, although idiosyncrasies may be observed due to biomechanical disposition or pathology. To better represent datasets with local-global structure, this work extends Convolutional Dictionary Learning (CDL), a popular method for learning interpretable representations, or dictionaries, of time-series data. In particular, we propose Personalized CDL (PerCDL), in which a local dictionary models local information as a personalized spatiotemporal transformation of a global dictionary. The transformation is learnable and can combine operations such as time warping and rotation. Formal computational and statistical guarantees for PerCDL are provided and its effectiveness on synthetic and real human locomotion data is demonstrated.

Network reconstruction is the task of inferring the unseen interactions between elements of a system, based only on their behavior or dynamics. This inverse problem is in general ill-posed, and admits many solutions for the same observation. Nevertheless, the vast majority of statistical methods proposed for this task -- formulated as the inference of a graphical generative model -- can only produce a ``point estimate,'' i.e. a single network considered the most likely. In general, this can give only a limited characterization of the reconstruction, since uncertainties and competing answers cannot be conveyed, even if their probabilities are comparable, while being structurally different. In this work we present an efficient MCMC algorithm for sampling from posterior distributions of reconstructed networks, which is able to reveal the full population of answers for a given reconstruction problem, weighted according to their plausibilities. Our algorithm is general, since it does not rely on specific properties of particular generative models, and is specially suited for the inference of large and sparse networks, since in this case an iteration can be performed in time $O(N\log^2 N)$ for a network of $N$ nodes, instead of $O(N^2)$, as would be the case for a more naive approach. We demonstrate the suitability of our method in providing uncertainties and consensus of solutions (which provably increases the reconstruction accuracy) in a variety of synthetic and empirical cases.

Basketball analytics has significantly advanced our understanding of the game, with shot selection emerging as a critical factor in both individual and team performance. With the advent of player tracking technologies, a wealth of granular data on shot attempts has become available, enabling a deeper analysis of shooting behavior. However, modeling shot selection presents unique challenges due to the spatial and contextual complexities influencing shooting decisions. This paper introduces a novel approach to the analysis of basketball shot data, focusing on the spatial distribution of shot attempts, also known as intensity surfaces. We model these intensity surfaces using a Functional Bayesian Additive Regression Trees (FBART) framework, which allows for flexible, nonparametric regression, and uncertainty quantification while addressing the nonlinearity and nonstationarity inherent in shot selection patterns to provide a more accurate representation of the factors driving player performance; we further propose the Adaptive Functional Bayesian Additive Regression Trees (AFBART) model, which builds on FBART by introducing adaptive basis functions for improved computational efficiency and model fit. AFBART is particularly well suited for the analysis of two-dimensional shot intensity surfaces and provides a robust tool for uncovering latent patterns in shooting behavior. Through simulation studies and real-world applications to NBA player data, we demonstrate the effectiveness of the model in quantifying shooting tendencies, improving performance predictions, and informing strategic decisions for coaches, players, and team managers. This work represents a significant step forward in the statistical modeling of basketball shot selection and its applications in optimizing game strategies.

National Highway Traffic Safety Administration reported 7,345 pedestrian fatalities in the United States in 2022, making pedestrian safety a pressing issue in urban mobility. This study presents a novel probabilistic simulation framework integrating dynamic pedestrian crossing models and Monte Carlo simulations to evaluate safety under varying traffic conditions. The framework captures key influences on pedestrian decisions, such as traffic light states, vehicle proximity, and waiting times, while employing the Intelligent Driver Model (IDM) to simulate realistic vehicle dynamics. Results from 500 trials show that pedestrians avoid crossing during green lights, reducing collision risks, while shorter waiting times during red lights encourage safer crossings. The risk is heightened during yellow lights, especially with nearby vehicles. This research emphasizes the importance of adaptive traffic control measures, such as pedestrian-triggered signals and enhanced traffic light timing, to mitigate risks and prioritize pedestrian safety. By modeling realistic interactions between pedestrians and vehicles, the study offers insights for designing safer and more sustainable urban intersections.

The theory of optimal transportation has developed into a powerful and elegant framework for comparing probability distributions, with wide-ranging applications in all areas of science. The fundamental idea of analyzing probabilities by comparing their underlying state space naturally aligns with the core idea of causal inference, where understanding and quantifying counterfactual states is paramount. Despite this intuitive connection, explicit research at the intersection of optimal transport and causal inference is only beginning to develop. Yet, many foundational models in causal inference have implicitly relied on optimal transport principles for decades, without recognizing the underlying connection. Therefore, the goal of this review is to offer an introduction to the surprisingly deep existing connections between optimal transport and the identification of causal effects with observational data -- where optimal transport is not just a set of potential tools, but actually builds the foundation of model assumptions. As a result, this review is intended to unify the language and notation between different areas of statistics, mathematics, and econometrics, by pointing out these existing connections, and to explore novel problems and directions for future work in both areas derived from this realization.

We study the sample complexity of pure exploration in an online learning problem with a feedback graph. This graph dictates the feedback available to the learner, covering scenarios between full-information, pure bandit feedback, and settings with no feedback on the chosen action. While variants of this problem have been investigated for regret minimization, no prior work has addressed the pure exploration setting, which is the focus of our study. We derive an instance-specific lower bound on the sample complexity of learning the best action with fixed confidence, even when the feedback graph is unknown and stochastic, and present unidentifiability results for Bernoulli rewards. Additionally, our findings reveal how the sample complexity scales with key graph-dependent quantities. Lastly, we introduce TaS-FG (Track and Stop for Feedback Graphs), an asymptotically optimal algorithm, and demonstrate its efficiency across different graph configurations.

Recent developments in big data analysis, machine learning, Industry 4.0, and IoT applications have enabled the monitoring and processing of multi-sensor data collected from systems, allowing for the prediction of the "Remaining Useful Life" (RUL) of system components. Particularly in the aviation industry, Prognostic Health Management (PHM) has become one of the most important practices for ensuring reliability and safety. Not only is the accuracy of RUL prediction important, but the implementability of techniques, domain adaptability, and interpretability of system degradation behaviors have also become essential. In this paper, the data collected from the multi-sensor environment of complex systems are processed using a Functional Data Analysis (FDA) approach to predict when the systems will fail and to understand and interpret the systems' life cycles. The approach is applied to the C-MAPSS datasets shared by National Aeronautics and Space Administration, and the behaviors of the sensors in aircraft engine failures are adaptively modeled with Multivariate Functional Principal Component Analysis (MFPCA). While the results indicate that the proposed method predicts the RUL competitively compared to other methods in the literature, it also demonstrates how multivariate Functional Data Analysis is useful for interpretability in prognostic studies within multi-sensor environments.

Pure exploration is one of the fundamental problems in multi-armed bandits (MAB). However, existing works mostly focus on specific pure exploration tasks, without a holistic view of the general pure exploration problem. This work fills this gap by introducing a versatile framework to study pure exploration, with a focus on identifying the pairwise relationships between targeted arm pairs. Moreover, unlike existing works that only optimize the stopping time (i.e., sample complexity), this work considers that arms are associated with potentially different costs and targets at optimizing the cumulative cost that occurred during learning. Under the general framework of pairwise pure exploration with arm-specific costs, a performance lower bound is derived. Then, a novel algorithm, termed CAET (Cost-Aware Pairwise Exploration Task), is proposed. CAET builds on the track-and-stop principle with a novel design to handle the arm-specific costs, which can potentially be zero and thus represent a very challenging case. Theoretical analyses prove that the performance of CAET approaches the lower bound asymptotically. Special cases are further discussed, including an extension to regret minimization, which is another major focus of MAB. The effectiveness and efficiency of CAET are also verified through experimental results under various settings.

COVID-19 has had a large scale negative impact on the health of opioid users exacerbating the health of an already vulnerable population. Critical information on the total impact of COVID-19 on opioid users is unknown due to a lack of comprehensive data on COVID-19 cases, inaccurate diagnostic coding, and lack of data coverage. To assess the impact of COVID-19 on small-area opioid mortality, we developed a Bayesian hierarchical excess opioid mortality modeling approach. We incorporate spatio-temporal autocorrelation structures to allow for sharing of information across small areas and time to reduce uncertainty in small area estimates. Excess mortality is defined as the difference between observed trends after a crisis and expected trends based on observed historical trends, which captures the total increase in observed mortality rates compared to what was expected prior to the crisis. We illustrate the application of our approach to assess excess opioid mortality risk estimates for 159 counties in GA. Using our proposed approach will help inform interventions in opioid-related public health responses, policies, and resource allocation. The application of this work also provides a general framework for improving the estimation and mapping of health indicators during crisis periods for the opioid user population.

For a given base class of sequence-to-next-token generators, we consider learning prompt-to-answer mappings obtained by iterating a fixed, time-invariant generator for multiple steps, thus generating a chain-of-thought, and then taking the final token as the answer. We formalize the learning problems both when the chain-of-thought is observed and when training only on prompt-answer pairs, with the chain-of-thought latent. We analyze the sample and computational complexity both in terms of general properties of the base class (e.g. its VC dimension) and for specific base classes such as linear thresholds. We present a simple base class that allows for universal representability and computationally tractable chain-of-thought learning. Central to our development is that time invariance allows for sample complexity that is independent of the length of the chain-of-thought. Attention arises naturally in our construction.

In this work, we investigate the behavior of ridge regression in an overparameterized binary classification task. We assume examples are drawn from (anisotropic) class-conditional cluster distributions with opposing means and we allow for the training labels to have a constant level of label-flipping noise. We characterize the classification error achieved by ridge regression under the assumption that the covariance matrix of the cluster distribution has a high effective rank in the tail. We show that ridge regression has qualitatively different behavior depending on the scale of the cluster mean vector and its interaction with the covariance matrix of the cluster distributions. In regimes where the scale is very large, the conditions that allow for benign overfitting turn out to be the same as those for the regression task. We additionally provide insights into how the introduction of label noise affects the behavior of the minimum norm interpolator (MNI). The optimal classifier in this setting is a linear transformation of the cluster mean vector and in the noiseless setting the MNI approximately learns this transformation. On the other hand, the introduction of label noise can significantly change the geometry of the solution while preserving the same qualitative behavior.

This paper investigates approximation capabilities of two-dimensional (2D) deep convolutional neural networks (CNNs), with Korobov functions serving as a benchmark. We focus on 2D CNNs, comprising multi-channel convolutional layers with zero-padding and ReLU activations, followed by a fully connected layer. We propose a fully constructive approach for building 2D CNNs to approximate Korobov functions and provide rigorous analysis of the complexity of the constructed networks. Our results demonstrate that 2D CNNs achieve near-optimal approximation rates under the continuous weight selection model, significantly alleviating the curse of dimensionality. This work provides a solid theoretical foundation for 2D CNNs and illustrates their potential for broader applications in function approximation.

Denoising diffusions provide a general strategy to sample from a probability distribution $\mu$ in $\mathbb{R}^d$ by constructing a stochastic process $(\hat{\boldsymbol x}_t:t\ge 0)$ in ${\mathbb R}^d$ such that the distribution of $\hat{\boldsymbol x}_T$ at large times $T$ approximates $\mu$. The drift ${\boldsymbol m}:{\mathbb R}^d\times{\mathbb R}\to{\mathbb R}^d$ of this diffusion process is learned from data (samples from $\mu$) by minimizing the so-called score-matching objective. In order for the generating process to be efficient, it must be possible to evaluate (an approximation of) ${\boldsymbol m}({\boldsymbol y},t)$ in polynomial time. Is every probability distribution $\mu$, for which sampling is tractable, also amenable to sampling via diffusions? We provide evidence to the contrary by constructing a probability distribution $\mu$ for which sampling is easy, but the drift of the diffusion process is intractable -- under a popular conjecture on information-computation gaps in statistical estimation. We further show that any polynomial-time computable drift can be modified in a way that changes minimally the score matching objective and yet results in incorrect sampling.

Motivated by privacy concerns in sequential decision-making on sensitive data, we address the challenge of nonparametric contextual multi-armed bandits (MAB) under local differential privacy (LDP). We develop a uniform-confidence-bound-type estimator, showing its minimax optimality supported by a matching minimax lower bound. We further consider the case where auxiliary datasets are available, subject also to (possibly heterogeneous) LDP constraints. Under the widely-used covariate shift framework, we propose a jump-start scheme to effectively utilize the auxiliary data, the minimax optimality of which is further established by a matching lower bound. Comprehensive experiments on both synthetic and real-world datasets validate our theoretical results and underscore the effectiveness of the proposed methods.

This study proposes median consensus embedding (MCE) to address variability in low-dimensional embeddings caused by random initialization in dimensionality reduction techniques such as t-distributed stochastic neighbor embedding. MCE is defined as the geometric median of multiple embeddings. By assuming multiple embeddings as independent and identically distributed random samples and applying large deviation theory, we prove that MCE achieves consistency at an exponential rate. Furthermore, we develop a practical algorithm to implement MCE by constructing a distance function between embeddings based on the Frobenius norm of the pairwise distance matrix of data points. Application to real-world data demonstrates that MCE converges rapidly and significantly reduces instability. These results confirm that MCE effectively mitigates random initialization issues in embedding methods.

Empirical substantive research, such as in the life or social sciences, is commonly categorized into the two modes exploratory and confirmatory, both of which are essential to scientific progress. The former is also referred to as hypothesis-generating or data-contingent research, the latter is also called hypothesis-testing research. In the context of empirical methodological research in statistics, however, the exploratory-confirmatory distinction has received very little attention so far. Our paper aims to fill this gap. First, we revisit the concept of empirical methodological research through the lens of the exploratory-confirmatory distinction. Secondly, we examine current practice with respect to this distinction through a literature survey including 115 articles from the field of biostatistics. Thirdly, we provide practical recommendations towards more appropriate design, interpretation, and reporting of empirical methodological research in light of this distinction. In particular, we argue that both modes of research are crucial to methodological progress, but that most published studies - even if sometimes disguised as confirmatory - are essentially of exploratory nature. We emphasize that it may be adequate to consider empirical methodological research as a continuum between "pure" exploration and "strict" confirmation, recommend to transparently report the mode of conducted research within the spectrum between exploratory and confirmatory, and stress the importance of study protocols written before conducting the study, especially in confirmatory methodological research.

We discuss necessary conditions for a PAC-Bayes bound to provide a meaningful generalisation guarantee. Our analysis reveals that the optimal generalisation guarantee depends solely on the distribution of the risk induced by the prior distribution. In particular, achieving a target generalisation level is only achievable if the prior places sufficient mass on high-performing predictors. We relate these requirements to the prevalent practice of using data-dependent priors in deep learning PAC-Bayes applications, and discuss the implications for the claim that PAC-Bayes ``explains'' generalisation.

Environmental and mental conditions are known risk factors for dermatitis and symptoms of skin inflammation, but their interplay is difficult to quantify; epidemiological studies rarely include both, along with possible confounding factors. Infodemiology leverages large online data sets to address this issue, but fusing them produces strong patterns of spatial and temporal correlation, missingness, and heterogeneity. In this paper, we design a causal network that correctly models these complex structures in large-scale infodemiological data from Google, EPA, NOAA and US Census (434 US counties, 134 weeks). Our model successfully captures known causal relationships between weather patterns, pollutants, mental conditions, and dermatitis. Key findings reveal that anxiety accounts for 57.4% of explained variance in dermatitis, followed by NO2 (33.9%), while environmental factors show significant mediation effects through mental conditions. The model predicts that reducing PM2.5 emissions by 25% could decrease dermatitis prevalence by 18%. Through statistical validation and causal inference, we provide unprecedented insights into the complex interplay between environmental and mental health factors affecting dermatitis, offering valuable guidance for public health policies and environmental regulations.

Bayesian inference for hierarchical models can be very challenging. MCMC methods have difficulty scaling to large models with many observations and latent variables. While variational inference (VI) and reweighted wake-sleep (RWS) can be more scalable, they are gradient-based methods and so often require many iterations to converge. Our key insight was that modern massively parallel importance weighting methods (Bowyer et al., 2024) give fast and accurate posterior moment estimates, and we can use these moment estimates to rapidly learn an approximate posterior. Specifically, we propose using expectation maximization to fit the approximate posterior, which we call QEM. The expectation step involves computing the posterior moments using high-quality massively parallel estimates from Bowyer et al. (2024). The maximization step involves fitting the approximate posterior using these moments, which can be done straightforwardly for simple approximate posteriors such as Gaussian, Gamma, Beta, Dirichlet, Binomial, Multinomial, Categorical, etc. (or combinations thereof). We show that QEM is faster than state-of-the-art, massively parallel variants of RWS and VI, and is invariant to reparameterizations of the model that dramatically slow down gradient based methods.

This work addresses the problem of estimating a vector field from a noisy Ordinary Differential Equation (ODE) in a non-parametric regression setting with a random design for initial values. More specifically, given a vector field $ f:\mathbb{R}^{D}\rightarrow \mathbb{R}^{D}$ governing a dynamical system defined by the autonomous ODE: $y' = f(y)$, we assume that the observations are $\tilde{y}_{X_{i}}(t_{j}) = y_{X_{i}}(t_{j}) + \varepsilon_{i,j}$ where $y_{X_{i}}(t_{j})$ is the solution of the ODE at time $t_{j}$ with initial condition $y(0) = X_{i}$, $X_{i}$ is sampled from a probability distribution $\mu$, and $\varepsilon_{i,j}$ some noise. In this context, we investigate, from a minimax perspective, the pointwise reconstruction of $f$ within the envelope of trajectories originating from the support of $\mu$. We propose an estimation strategy based on preliminary flow reconstruction and techniques from derivative estimation in non-parametric regression. Under mild assumptions on $f$, we establish convergence rates that depend on the temporal resolution, the number of sampled initial values and the mass concentration of $\mu$. Importantly, we show that these rates are minimax optimal. Furthermore, we discuss the implications of our results in a manifold learning setting, providing insights into how our approach can mitigate the curse of dimensionality.

Motivated by the need to analyze continuously updated data sets in the context of time-to-event modeling, we propose a novel nonparametric approach to estimate the conditional hazard function given a set of continuous and discrete predictors. The method is based on a representation of the conditional hazard as a ratio between a joint density and a conditional expectation determined by the distribution of the observed variables. It is shown that such ratio representations are available for uni- and bivariate time-to-events, in the presence of common types of random censoring, truncation, and with possibly cured individuals, as well as for competing risks. This opens the door to nonparametric approaches in many time-to-event predictive models. To estimate joint densities and conditional expectations we propose the recursive kernel smoothing, which is well suited for online estimation. Asymptotic results for such estimators are derived and it is shown that they achieve optimal convergence rates. Simulation experiments show the good finite sample performance of our recursive estimator with right censoring. The method is applied to a real dataset of primary breast cancer.

Objectives: To examine the distribution, temporal associations, and age/sex-specific patterns of multiple long-term conditions (MLTCs) in adults with intellectual disability (ID). Study Design: Observational study using longitudinal healthcare data. Methods: Analysis of 18144 adults with ID (10168 males and 7976 females) identified in the Clinical Practice Research Datalink, linked to Hospital Episode Statistics Admitted Patient Care and Outpatient data (2000-2021). We used temporal analysis to establish directional associations among 40 long-term conditions, stratified by sex and age groups (under 45, 45-64, 65 and over). Results: The high prevalence of enduring mental illness across all age groups is an important finding unique to this population. In males, mental illness occurred along with upper gastrointestinal conditions (specifically reflux disorders), while in females, mental illness presented alongside reflux disorders, chronic pain, and endocrine conditions such as thyroid problems. Among young males with intellectual disability, the combination of cerebral palsy with dysphagia, epilepsy, chronic constipation, and chronic pneumonia represents a distinctive pattern. In those aged 45-64, we observed early onset of lifestyle conditions like diabetes and hypertension, though notably these conditions co-occurred with mental illness and anaemia at rates exceeding those in the general population. The health conditions in those aged 65 and over largely mirrored those seen in the general aging population. Conclusions: Our findings highlight the complex patterns of MLTCs in this population, revealing sex-specific associations across age groups, and identified temporal associations, thus providing insights into disease progression, which can inform targeted prevention strategies and interventions to prevent premature mortality.

Evaluation of clinical prediction models across multiple clusters, whether centers or datasets, is becoming increasingly common. A comprehensive evaluation includes an assessment of the agreement between the estimated risks and the observed outcomes, also known as calibration. Calibration is of utmost importance for clinical decision making with prediction models and it may vary between clusters. We present three approaches to take clustering into account when evaluating calibration. (1) Clustered group calibration (CG-C), (2) two stage meta-analysis calibration (2MA-C) and (3) mixed model calibration (MIX-C) can obtain flexible calibration plots with random effects modelling and providing confidence and prediction intervals. As a case example, we externally validate a model to estimate the risk that an ovarian tumor is malignant in multiple centers (N = 2489). We also conduct a simulation study and synthetic data study generated from a true clustered dataset to evaluate the methods. In the simulation and the synthetic data analysis MIX-C gave estimated curves closest to the true overall and center specific curves. Prediction interval was best for 2MA-C with splines. Standard flexible calibration worked likewise in terms of calibration error when sample size is limited. We recommend using 2MA-C (splines) to estimate the curve with the average effect and the 95% PI and MIX-C for the cluster specific curves, specially when sample size per cluster is limited. We provide ready-to-use code to construct summary flexible calibration curves with confidence and prediction intervals to assess heterogeneity in calibration across datasets or centers.

Early work established convergence of the principal component estimators of the factors and loadings up to a rotation for large dimensional approximate factor models with weak factors in that the factor loading $\Lambda^{(0)}$ scales sublinearly in the number $N$ of cross-section units, i.e., $\Lambda^{(0)\top}\Lambda^{(0)}/N^{\alpha}$ is positive definite in the limit for some $\alpha\in (0,1)$. However, the established convergence rates for weak factors can be much slower especially for small $\alpha$. This article proposes a Transfer Principal Component Analysis (TransPCA) method for enhancing the convergence rates for weak factors by transferring knowledge from large number of available informative panel datasets, which should not be turned a blind eye on in this big data era. We aggregate useful information by analyzing a weighted average projection matrix of the estimated loading spaces from all informative datasets which is highly flexible and computationally efficient. Theoretically, we derive the convergence rates of the estimators of weak/strong loading spaces and factor scores. The results indicate that as long as the auxiliary datasets are similar enough to the target dataset and the auxiliary sample size is sufficiently large, TransPCA estimators can achieve faster convergence rates in contrast to performing PCA solely on the target dataset. To avoid negative transfer, we also investigate the case that the informative datasets are unknown and provide a criterion for selecting useful datasets. Thorough simulation studies and {empirical analysis on real datasets in areas of macroeconomic and finance} are conducted to illustrate the usefulness of our proposed methods where large number of source panel datasets are naturally available.

The information criterion AIC has been used successfully in many areas of statistical modeling, and since it is derived based on the Taylor expansion of the log-likelihood function and the asymptotic distribution of the maximum likelihood estimator, it is not directly justified for likelihood functions that include non-differentiable points such as the Laplace distribution. In fact, it is known to work effectively in many such cases. In this paper, we attempt to evaluate the bias correction directly for the case where the true model or the model to be estimated is a simple Laplace distribution model. As a result, an approximate expression for the bias correction term was obtained. Numerical results show that the AIC approximations are relatively good except when the Gauss distribution model is fitted to data following the Laplace distribution.

Jointly modeling and forecasting economic and financial variables across a large set of countries has long been a significant challenge. Two primary approaches have been utilized to address this issue: the vector autoregressive model with exogenous variables (VARX) and the matrix autoregression (MAR). The VARX model captures domestic dependencies, but treats variables exogenous to represent global factors driven by international trade. In contrast, the MAR model simultaneously considers variables from multiple countries but ignores the trade network. In this paper, we propose an extension of the MAR model that achieves these two aims at once, i.e., studying both international dependencies and the impact of the trade network on the global economy. Additionally, we introduce a sparse component to the model to differentiate between systematic and idiosyncratic cross-predictability. To estimate the model parameters, we propose both a likelihood estimation method and a bias-corrected alternating minimization version. We provide theoretical and empirical analyses of the model's properties, alongside presenting intriguing economic insights derived from our findings.

Survival analysis plays a crucial role in understanding time-to-event (survival) outcomes such as disease progression. Despite recent advancements in causal mediation frameworks for survival analysis, existing methods are typically based on Cox regression and primarily focus on a single exposure or individual omics layers, often overlooking multi-omics interplay. This limitation hinders the full potential of integrated biological insights. In this paper, we propose SMAHP, a novel method for survival mediation analysis that simultaneously handles high-dimensional exposures and mediators, integrates multi-omics data, and offers a robust statistical framework for identifying causal pathways on survival outcomes. This is one of the first attempts to introduce the accelerated failure time (AFT) model within a multi-omics causal mediation framework for survival outcomes. Through simulations across multiple scenarios, we demonstrate that SMAHP achieves high statistical power, while effectively controlling false discovery rate (FDR), compared with two other approaches. We further apply SMAHP to the largest head-and-neck carcinoma proteogenomic data, detecting a gene mediated by a protein that influences survival time.

Curse of Dimensionality is an unavoidable challenge in statistical probability models, yet diffusion models seem to overcome this limitation, achieving impressive results in high-dimensional data generation. Diffusion models assume that they can learn the statistical properties of the underlying probability distribution, enabling sampling from this distribution to generate realistic samples. But is this really how they work? To address this question, this paper conducts a detailed analysis of the objective function and inference methods of diffusion models, leading to several important conclusions that help answer the above question: 1) In high-dimensional sparse scenarios, the target of the objective function fitting degrades from a weighted sum of multiple samples to a single sample. 2) The mainstream inference methods can all be represented within a simple unified framework, without requiring statistical concepts such as Markov chains and SDEs. 3) Guided by this simple framework, more efficient inference methods can be discovered.

National forest inventory (NFI) data are often costly to collect, which inhibits efforts to estimate parameters of interest for small spatial, temporal, or biophysical domains. Traditionally, design-based estimators are used to estimate status of forest parameters of interest, but are unreliable for small areas where data are sparse. Additionally, design-based estimates constructed directly from the survey data are often unavailable when sample sizes are especially small. Traditional model-based small area estimation approaches, such as the Fay-Herriot (FH) model, rely on these direct estimates for inference; hence, missing direct estimates preclude the use of such approaches. Here, we detail a Bayesian spatio-temporal small area estimation model that efficiently leverages sparse NFI data to estimate status and trends for forest parameters. The proposed model bypasses the use of direct estimates and instead uses plot-level NFI measurements along with auxiliary data including remotely sensed tree canopy cover. We produce forest carbon estimates from the United States NFI over 14 years across the contiguous US (CONUS) and conduct a simulation study to assess our proposed model's accuracy, precision, and bias, compared to that of a design-based estimator. The proposed model provides improved precision and accuracy over traditional estimation methods, and provides useful insights into county-level forest carbon dynamics across the CONUS.

In this article, we propose a novel logistic quasi-maximum likelihood estimation (LQMLE) for general parametric time series models. Compared to the classical Gaussian QMLE and existing robust estimations, it enjoys many distinctive advantages, such as robustness in respect of distributional misspecification and heavy-tailedness of the innovation, more resiliency to outliers, smoothness and strict concavity of the log logistic quasi-likelihood function, and boundedness of the influence function among others. Under some mild conditions, we establish the strong consistency and asymptotic normality of the LQMLE. Moreover, we propose a new and vital parameter identifiability condition to ensure desirable asymptotics of the LQMLE. Further, based on the LQMLE, we consider the Wald test and the Lagrange multiplier test for the unknown parameters, and derive the limiting distributions of the corresponding test statistics. The applicability of our methodology is demonstrated by several time series models, including DAR, GARCH, ARMA-GARCH, DTARMACH, and EXPAR. Numerical simulation studies are carried out to assess the finite-sample performance of our methodology, and an empirical example is analyzed to illustrate its usefulness.

Additive models offer accurate and interpretable predictions for tabular data, a critical tool for statistical modeling. Recent advances in Neural Additive Models (NAMs) allow these models to handle complex machine learning tasks, including feature selection and survival analysis, on large-scale data. This paper introduces dnamite, a Python package that implements NAMs for these advanced applications. dnamite provides a scikit-learn style interface to train regression, classification, and survival analysis NAMs, with built-in support for feature selection. We describe the methodology underlying dnamite, its design principles, and its implementation. Through an application to the MIMIC III clinical dataset, we demonstrate the utility of dnamite in a real-world setting where feature selection and survival analysis are both important. The package is publicly available via pip and documented at dnamite.readthedocs.io.

The Antibiotic Resistance Microbiology Dataset (ARMD) is a de-identified resource derived from electronic health records (EHR) that facilitates research into antimicrobial resistance (AMR). ARMD encompasses data from adult patients, focusing on microbiological cultures, antibiotic susceptibilities, and associated clinical and demographic features. Key attributes include organism identification, susceptibility patterns for 55 antibiotics, implied susceptibility rules, and de-identified patient information. This dataset supports studies on antimicrobial stewardship, causal inference, and clinical decision-making. ARMD is designed to be reusable and interoperable, promoting collaboration and innovation in combating AMR. This paper describes the dataset's acquisition, structure, and utility while detailing its de-identification process.

We study the classical and parameterized complexity of computing the positive non-clashing teaching dimension of a set of concepts, that is, the smallest number of examples per concept required to successfully teach an intelligent learner under the considered, previously established model. For any class of concepts, it is known that this problem can be effortlessly transferred to the setting of balls in a graph G. We establish (1) the NP-hardness of the problem even when restricted to instances with positive non-clashing teaching dimension k=2 and where all balls in the graph are present, (2) near-tight running time upper and lower bounds for the problem on general graphs, (3) fixed-parameter tractability when parameterized by the vertex integrity of G, and (4) a lower bound excluding fixed-parameter tractability when parameterized by the feedback vertex number and pathwidth of G, even when combined with k. Our results provide a nearly complete understanding of the complexity landscape of computing the positive non-clashing teaching dimension and answer open questions from the literature.

Resource-efficiently computing representations of probability distributions and the distances between them while only having access to the samples is a fundamental and useful problem across mathematical sciences. In this paper, we propose a generic algorithmic framework to estimate the PDF and CDF of any sub-Gaussian distribution while the samples from them arrive in a stream. We compute mergeable summaries of distributions from the stream of samples that require sublinear space w.r.t. the number of observed samples. This allows us to estimate Wasserstein and Total Variation (TV) distances between any two sub-Gaussian distributions while samples arrive in streams and from multiple sources (e.g. federated learning). Our algorithms significantly improves on the existing methods for distance estimation incurring super-linear time and linear space complexities. In addition, we use the proposed estimators of Wasserstein and TV distances to audit the fairness and privacy of the ML algorithms. We empirically demonstrate the efficiency of the algorithms for estimating these distances and auditing using both synthetic and real-world datasets.

We present a semi-supervised fine-tuning framework for foundation models that utilises mutual information decomposition to address the challenges of training for a limited amount of labelled data. Our approach derives two distinct lower bounds: i) for the downstream task space, such as classification, optimised using conditional and marginal cross-entropy alongside Kullback-Leibler divergence, and ii) for the latent space representation, regularised and aligned using a contrastive-like decomposition. This fine-tuning strategy retains the pre-trained structure of the foundation model, modifying only a specialised projector module comprising a small transformer and a token aggregation technique. Experiments on several datasets demonstrate significant improvements in classification tasks under extremely low-labelled conditions by effectively leveraging unlabelled data.

Counterfactual explanations indicate the smallest change in input that can translate to a different outcome for a machine learning model. Counterfactuals have generated immense interest in high-stakes applications such as finance, education, hiring, etc. In several use-cases, the decision-making process often relies on an ensemble of models rather than just one. Despite significant research on counterfactuals for one model, the problem of generating a single counterfactual explanation for an ensemble of models has received limited interest. Each individual model might lead to a different counterfactual, whereas trying to find a counterfactual accepted by all models might significantly increase cost (effort). We propose a novel strategy to find the counterfactual for an ensemble of models using the perspective of entropic risk measure. Entropic risk is a convex risk measure that satisfies several desirable properties. We incorporate our proposed risk measure into a novel constrained optimization to generate counterfactuals for ensembles that stay valid for several models. The main significance of our measure is that it provides a knob that allows for the generation of counterfactuals that stay valid under an adjustable fraction of the models. We also show that a limiting case of our entropic-risk-based strategy yields a counterfactual valid for all models in the ensemble (worst-case min-max approach). We study the trade-off between the cost (effort) for the counterfactual and its validity for an ensemble by varying degrees of risk aversion, as determined by our risk parameter knob. We validate our performance on real-world datasets.

While deep generative models have significantly advanced representation learning, they may inherit or amplify biases and fairness issues by encoding sensitive attributes alongside predictive features. Enforcing strict independence in disentanglement is often unrealistic when target and sensitive factors are naturally correlated. To address this challenge, we propose CAD-VAE (Correlation-Aware Disentangled VAE), which introduces a correlated latent code to capture the shared information between target and sensitive attributes. Given this correlated latent, our method effectively separates overlapping factors without extra domain knowledge by directly minimizing the conditional mutual information between target and sensitive codes. A relevance-driven optimization strategy refines the correlated code by efficiently capturing essential correlated features and eliminating redundancy. Extensive experiments on benchmark datasets demonstrate that CAD-VAE produces fairer representations, realistic counterfactuals, and improved fairness-aware image editing.

In this paper, we propose a method for density-based clustering in high-dimensional spaces that combines Locality-Sensitive Hashing (LSH) with the Quick Shift algorithm. The Quick Shift algorithm, known for its hierarchical clustering capabilities, is extended by integrating approximate Kernel Density Estimation (KDE) using LSH to provide efficient density estimates. The proposed approach achieves almost linear time complexity while preserving the consistency of density-based clustering.

In recent years the information asymmetric Lipschitz bandits In this paper we studied the Lipschitz bandit problem applied to the multiplayer information asymmetric problem studied in \cite{chang2022online, chang2023optimal}. More specifically we consider information asymmetry in rewards, actions, or both. We adopt the CAB algorithm given in \cite{kleinberg2004nearly} which uses a fixed discretization to give regret bounds of the same order (in the dimension of the action) space in all 3 problem settings. We also adopt their zooming algorithm \cite{ kleinberg2008multi}which uses an adaptive discretization and apply it to information asymmetry in rewards and information asymmetry in actions.

Graph sparsification is a well-established technique for accelerating graph-based learning algorithms, which uses edge sampling to approximate dense graphs with sparse ones. Because the sparsification error is random and unknown, users must contend with uncertainty about the reliability of downstream computations. Although it is possible for users to obtain conceptual guidance from theoretical error bounds in the literature, such results are typically impractical at a numerical level. Taking an alternative approach, we propose to address these issues from a data-driven perspective by computing empirical error estimates. The proposed error estimates are highly versatile, and we demonstrate this in four use cases: Laplacian matrix approximation, graph cut queries, graph-structured regression, and spectral clustering. Moreover, we provide two theoretical guarantees for the error estimates, and explain why the cost of computing them is manageable in comparison to the overall cost of a typical graph sparsification workflow.

This paper examines the evolution of the Finnish electric energy system up to 2035, focusing on the likelihood of different development paths. The primary contribution of this paper is the development of an extensive Bayesian Network, designed to model and analyse the evolution of power generation capacity mix, assess the likelihood of different grid management scenarios, and understand the causal relationships underlying these scenarios. A target optimisation was carried out using the constructed Bayesian Network to explore possibilities to minimise grid management complexity. The results of the optimisation reveal that the authorities and stakeholders should prioritise increasing demand response, gas power, and battery storage capacities. These mature technologies are well-suited to guarantee energy adequacy during peak consumption periods, which in Finland typically occur during consecutive cold, dark and windless winter weeks. Although this study focuses on the evolution of the Finnish power grid, the constructed Bayesian Network approach is broadly applicable and can be utilised to explore causal relationships in other countries by employing the designed questionnaire and engaging a panel of experts specific to the country's energy infrastructure.

Instead of performing text-conditioned denoising in the image domain, latent diffusion models (LDMs) operate in latent space of a variational autoencoder (VAE), enabling more efficient processing at reduced computational costs. However, while the diffusion process has moved to the latent space, the contrastive language-image pre-training (CLIP) models, as used in many image processing tasks, still operate in pixel space. Doing so requires costly VAE-decoding of latent images before they can be processed. In this paper, we introduce Latent-CLIP, a CLIP model that operates directly in the latent space. We train Latent-CLIP on 2.7B pairs of latent images and descriptive texts, and show that it matches zero-shot classification performance of similarly sized CLIP models on both the ImageNet benchmark and a LDM-generated version of it, demonstrating its effectiveness in assessing both real and generated content. Furthermore, we construct Latent-CLIP rewards for reward-based noise optimization (ReNO) and show that they match the performance of their CLIP counterparts on GenEval and T2I-CompBench while cutting the cost of the total pipeline by 21%. Finally, we use Latent-CLIP to guide generation away from harmful content, achieving strong performance on the inappropriate image prompts (I2P) benchmark and a custom evaluation, without ever requiring the costly step of decoding intermediate images.

The real-time process of directional changes while drilling, known as geosteering, is crucial for hydrocarbon extraction and emerging directional drilling applications such as geothermal energy, civil infrastructure, and CO2 storage. The geo-energy industry seeks an automatic geosteering workflow that continually updates the subsurface uncertainties and captures the latest geological understanding given the most recent observations in real-time. We propose "DISTINGUISH": a real-time, AI-driven workflow designed to transform geosteering by integrating Generative Adversarial Networks (GANs) for geological parameterization, ensemble methods for model updating, and global discrete dynamic programming (DDP) optimization for complex decision-making during directional drilling operations. The DISTINGUISH framework relies on offline training of a GAN model to reproduce relevant geology realizations and a Forward Neural Network (FNN) to model Logging-While-Drilling (LWD) tools' response for a given geomodel. This paper introduces a first-of-its-kind workflow that progressively reduces GAN-geomodel uncertainty around and ahead of the drilling bit and adjusts the well plan accordingly. The workflow automatically integrates real-time LWD data with a DDP-based decision support system, enhancing predictive models of geology ahead of drilling and leading to better steering decisions. We present a simple yet representative benchmark case and document the performance target achieved by the DISTINGUISH workflow prototype. This benchmark will be a foundation for future methodological advancements and workflow refinements.

We investigate the role of inertia in the asynchronous state of a disordered Kuramoto model. We extend an iterative simulation scheme to the case of the Kuramoto model with inertia in order to determine the self-consistent fluctuation statistics, specifically, the power spectra of network noise and single oscillators. Comparison with network simulations demonstrates that this works well whenever the system is in an asynchronous state. We also find an unexpected effect when varying the degree of inertia: the correlation time of the oscillators becomes minimal at an intermediate mass of the oscillators; correspondingly, the power spectra appear flatter and thus more similar to white noise around the same value of mass. We also find a similar effect for the Lyapunov spectra of the oscillators when the mass is varied.