Papers

Identifying and comparing topological features, particularly cycles, across different topological objects remains a fundamental challenge in persistent homology and topological data analysis. This work introduces a novel framework for constructing cycle communities through two complementary approaches. First, a dendrogram-based methodology leverages merge-tree algorithms to construct hierarchical representations of homology classes from persistence intervals. The Wasserstein distance on merge trees is introduced as a metric for comparing dendrograms, establishing connections to hierarchical clustering frameworks. Through simulation studies, the discriminative power of dendrogram representations for identifying cycle communities is demonstrated. Second, an extension of Stratified Gradient Sampling simultaneously learns multiple filter functions that yield cycle barycenter functions capable of faithfully reconstructing distinct sets of cycles. The set of cycles each filter function can reconstruct constitutes cycle communities that are non-overlapping and partition the space of all cycles. Together, these approaches transform the problem of cycle matching into both a hierarchical clustering and topological optimization framework, providing principled methods to identify similar topological structures both within and across groups of topological objects.

We consider statistical inference in high-dimensional regression problems under affine constraints on the parameter space. The theoretical study of this is motivated by the study of genetic determinants of diseases, such as diabetes, using external information from mediating protein expression levels. Specifically, we develop rigorous methods for estimating genetic effects on diabetes-related continuous outcomes when these associations are constrained based on external information about genetic determinants of proteins, and genetic relationships between proteins and the outcome of interest. In this regard, we discuss multiple candidate estimators and study their theoretical properties, sharp large sample optimality, and numerical qualities under a high-dimensional proportional asymptotic framework.

Graph Convolutional Networks (GCNs) have become a standard approach for semi-supervised node classification, yet practitioners lack clear guidance on when GCNs provide meaningful improvements over simpler baselines. We present a diagnostic study using the Amazon Computers co-purchase data to understand when and why GCNs help. Through systematic experiments with simulated label scarcity, feature ablation, and per-class analysis, we find that GCN performance depends critically on the interaction between graph homophily and feature quality. GCNs provide the largest gains under extreme label scarcity, where they leverage neighborhood structure to compensate for limited supervision. Surprisingly, GCNs can match their original performance even when node features are replaced with random noise, suggesting that structure alone carries sufficient signal on highly homophilous graphs. However, GCNs hurt performance when homophily is low and features are already strong, as noisy neighbors corrupt good predictions. Our quadrant analysis reveals that GCNs help in three of four conditions and only hurt when low homophily meets strong features. These findings offer practical guidance for practitioners deciding whether to adopt graph-based methods.

Structural changes and outliers often coexist, complicating statistical inference. This paper addresses the problem of testing for parameter changes in conditionally heteroscedastic time series models, particularly in the presence of outliers. To mitigate the impact of outliers, we introduce a two-step procedure comprising robust estimation and residual truncation. Based on this procedure, we propose a residual-based robust CUSUM test and its self-normalized counterpart. We derive the limiting null distributions of the proposed robust tests and establish their consistency. We also investigate the breakdown of the naive tests under additive and innovation outlier contamination and provide theoretical support for the robustness of the proposed tests. Simulation results demonstrate the strong robustness of our tests against outliers. Finally, we analyze Bitcoin data to illustrate the practical application.

In this paper, we derive the asymptotic distribution of the number of copies of a fixed graph $H$ in a random graph $G_n$ sampled from a sparse graphon model. Specifically, we provide a refined analysis that separates the contributions of edge randomness and vertex-label randomness, allowing us to identify distinct sparsity regimes in which each component dominates or both contribute jointly to the fluctuations. As a result, we establish asymptotic normality for the count of any fixed graph $H$ in $G_n$ across the entire range of sparsity (above the containment threshold for $H$ in $G_n$). These results provide a complete description of subgraph count fluctuations in sparse inhomogeneous networks, closing several gaps in the existing literature that were limited to specific motifs or suboptimal sparsity assumptions.

We develop a rigorous walk-forward validation framework for algorithmic trading designed to mitigate overfitting and lookahead bias. Our methodology combines interpretable hypothesis-driven signal generation with reinforcement learning and strict out-of-sample testing. The framework enforces strict information set discipline, employs rolling window validation across 34 independent test periods, maintains complete interpretability through natural language hypothesis explanations, and incorporates realistic transaction costs and position constraints. Validating five market microstructure patterns across 100 US equities from 2015 to 2024, the system yields modest annualized returns (0.55%, Sharpe ratio 0.33) with exceptional downside protection (maximum drawdown -2.76%) and market-neutral characteristics (beta = 0.058). Performance exhibits strong regime dependence, generating positive returns during high-volatility periods (0.60% quarterly, 2020-2024) while underperforming in stable markets (-0.16%, 2015-2019). We report statistically insignificant aggregate results (p-value 0.34) to demonstrate a reproducible, honest validation protocol that prioritizes interpretability and extends naturally to advanced hypothesis generators, including large language models. The key empirical finding reveals that daily OHLCV-based microstructure signals require elevated information arrival and trading activity to function effectively. The framework provides complete mathematical specifications and open-source implementation, establishing a template for rigorous trading system evaluation that addresses the reproducibility crisis in quantitative finance research. For researchers, practitioners, and regulators, this work demonstrates that interpretable algorithmic trading strategies can be rigorously validated without sacrificing transparency or regulatory compliance.

This study evaluates thresholds for removing singular values from singular value decomposition-based low-rank approximations of deep neural network weight matrices. Each weight matrix is modeled as the sum of signal and noise matrices. The low-rank approximation is obtained by removing noise-related singular values using a threshold based on random matrix theory. To assess the adequacy of this threshold, we propose an evaluation metric based on the cosine similarity between the singular vectors of the signal and original weight matrices. The proposed metric is used in numerical experiments to compare two threshold estimation methods.

Linear Autoencoders (LAEs) have shown strong performance in state-of-the-art recommender systems. However, this success remains largely empirical, with limited theoretical understanding. In this paper, we investigate the generalizability – a theoretical measure of model performance in statistical learning – of multivariate linear regression and LAEs. We first propose a PAC-Bayes bound for multivariate linear regression, extending the earlier bound for single-output linear regression by Shalaeva et al., and establish sufficient conditions for its convergence. We then show that LAEs, when evaluated under a relaxed mean squared error, can be interpreted as constrained multivariate linear regression models on bounded data, to which our bound adapts. Furthermore, we develop theoretical methods to improve the computational efficiency of optimizing the LAE bound, enabling its practical evaluation on large models and real-world datasets. Experimental results demonstrate that our bound is tight and correlates well with practical ranking metrics such as Recall@K and NDCG@K.

Neural population activity often exhibits regime-dependent non-stationarity in the form of switching dynamics. Learning accurate switching dynamical system models can reveal how behavior is encoded in neural activity. Existing switching approaches have primarily focused on learning models from a single neural modality, either continuous Gaussian signals or discrete Poisson signals. However, multiple neural modalities are often recorded simultaneously to measure different spatiotemporal scales of brain activity, and all these modalities can encode behavior. Moreover, regime labels are typically unavailable in training data, posing a significant challenge for learning models of regime-dependent switching dynamics. To address these challenges, we develop a novel unsupervised learning algorithm that learns the parameters of switching multiscale dynamical system models using only multiscale neural observations. We demonstrate our method using both simulations and two distinct experimental datasets with multimodal spike-LFP observations during different motor tasks. We find that our switching multiscale dynamical system models more accurately decode behavior than switching single-scale dynamical models, showing the success of multiscale neural fusion. Further, our models outperform stationary multiscale models, illustrating the importance of tracking regime-dependent non-stationarity in multimodal neural data. The developed unsupervised learning framework enables more accurate modeling of complex multiscale neural dynamics by leveraging information in multimodal recordings while incorporating regime switches. This approach holds promise for improving the performance and robustness of brain-computer interfaces over time and for advancing our understanding of the neural basis of behavior.

Electricity markets are under increasing pressure to maintain reliability amidst rising renewable penetration, demand variability, and occasional price shocks. Traditional capacity market designs often fall short in addressing this by relying on expected-value metrics of energy unserved, which overlook risk exposure in such systems. In this work, we present CapOptix, a capacity pricing framework that interprets capacity commitments as reliability options, i.e., financial derivatives of wholesale electricity prices. CapOptix characterizes the capacity premia charged by accounting for structural price shifts modeled by the Markov Regime Switching Process. We apply the framework to historical price data from multiple electricity markets and compare the resulting premium ranges with existing capacity remuneration mechanisms.

Bayesian hierarchical linear models provide a natural framework to analyze nested and clustered data. Classical estimation with Markov chain Monte Carlo produces well calibrated posterior distributions but becomes computationally expensive in high dimensional or large sample settings. Variational Inference and Stochastic Variational Inference offer faster optimization based alternatives, but their accuracy in hierarchical structures is uncertain when group separation is weak. This paper compares these two paradigms across three model classes, the Linear Regression Model, the Hierarchical Linear Regression Model, and a Clustered Hierarchical Linear Regression Model. Through simulation studies and an application to real data, the results show that variational methods recover global regression effects and clustering structure with a fraction of the computing time, but distort posterior dependence and yield unstable values of information criteria such as WAIC and DIC. The findings clarify when variational methods can serve as practical surrogates for Markov chain Monte Carlo and when their limitations make full Bayesian sampling necessary, and they provide guidance for extending the same variational framework to generalized linear models and other members of the exponential family.

Bayesian full waveform inversion (FWI) offers uncertainty-aware subsurface models; however, posterior sampling directly on observed seismic shot records is rarely practical at the field scale because each sample requires numerous wave-equation solves. We aim to make such sampling feasible for large surveys while preserving calibration, that is, high uncertainty in less illuminated areas. Our approach couples diffusion-based posterior sampling with simultaneous-source FWI data. At each diffusion noise level, a network predicts a clean velocity model. We then apply a stochastic refinement step in model space using Langevin dynamics under the wave-equation likelihood and reintroduce noise to decouple successive levels before proceeding. Simultaneous-source batches reduce forward and adjoint solves approximately in proportion to the supergather size, while an unconditional diffusion prior trained on velocity patches and volumes helps suppress source-related numerical artefacts. We evaluate the method on three 2D synthetic datasets (SEG/EAGE Overthrust, SEG/EAGE Salt, SEAM Arid), a 2D field line, and a 3D upscaling study. Relative to a particle-based variational baseline, namely Stein variational gradient descent without a learned prior and with single-source (non-simultaneous-source) FWI, our sampler achieves lower model error and better data fit at a substantially reduced computational cost. By aligning encoded-shot likelihoods with diffusion-based sampling and exploiting straightforward parallelization over samples and source batches, the method provides a practical path to calibrated posterior inference on observed shot records that scales to large 2D and 3D problems.

Clinical decision support tools built on electronic health records often experience performance drift due to temporal population shifts, particularly when changes in the clinical environment initially affect only a subset of patients, resulting in a transition to mixed populations. Such case-mix changes commonly arise following system-level operational updates or the emergence of new diseases, such as COVID-19. We propose TRACER (Transfer Learning-based Real-time Adaptation for Clinical Evolving Risk), a framework that identifies encounter-level transition membership and adapts predictive models using transfer learning without full retraining. In simulation studies, TRACER outperformed static models trained on historical or contemporary data. In a real-world application predicting hospital admission following emergency department visits across the COVID-19 transition, TRACER improved both discrimination and calibration. TRACER provides a scalable approach for maintaining robust predictive performance under evolving and heterogeneous clinical conditions.

Financial exclusion constrains entrepreneurship, increases income volatility, and widens wealth gaps. Underbanked consumers in Istanbul often have no bureau file because their earnings and payments flow through informal channels. To study how such borrowers can be evaluated we create a synthetic dataset of one hundred thousand Istanbul residents that reproduces first quarter 2025 TÜİK (TURKSTAT) census marginals and telecom usage patterns. Retrieval augmented generation feeds these public statistics into the OpenAI o3 model, which synthesises realistic yet private records. Each profile contains seven socio demographic variables and nine alternative attributes that describe phone specifications, online shopping rhythm, subscription spend, car ownership, monthly rent, and a credit card flag. To test the impact of the alternative financial data CatBoost, LightGBM, and XGBoost are each trained in two versions. Demo models use only the socio demographic variables; Full models include both socio demographic and alternative attributes. Across five fold stratified validation the alternative block raises area under the curve by about one point three percentage and lifts balanced F 1 from roughly 0.84 to 0.95, a fourteen percent gain. We contribute an open Istanbul 2025 Q1 synthetic dataset, a fully reproducible modeling pipeline, and empirical evidence that a concise set of behavioural attributes can approach bureau level discrimination power while serving borrowers who lack formal credit records. These findings give lenders and regulators a transparent blueprint for extending fair and safe credit access to the underbanked.

We study a class of robust mean estimators $\widehat{\mu}$ obtained by adaptively shrinking the weights of sample points far from a base estimator $\widehat{\kappa}$. Given a data-dependent scaling factor $\widehat{\alpha}$ and a weighting function $w:[0, \infty) \to [0,1]$, we let $\widehat{\mu}=\widehat{\kappa} + \frac{1}{n}\sum_{i=1}^n(X_i - \widehat{\kappa})w(\widehat{\alpha}|X_i-\widehat{\kappa}|)$. We prove that, under mild assumptions over $w$, these estimators achieve stronger concentration bounds than the base estimate $\widehat{\kappa}$, including sub-Gaussian guarantees. This framework unifies and extends several existing approaches to robust mean estimation in $\R$, and can also be generalized to the multivariate setting. Through numerical experiments, we show that our shrinking approach translates to faster concentration, even for small sample sizes.

Learning surrogate models for physical systems with latent uncertainty remains challenging in data-scarce regimes: deterministic neural operators fail to characterize uncertainty, while generative approaches require large ensembles of high-fidelity solution operator simulations and often sacrifice resolution generalizability. In this work, we propose a residual-augmented probabilistic operator learning framework that casts flow-matching-based generative modeling in infinite-dimensional function spaces while leveraging inexpensive low-fidelity solution operators as an inductive bias. Rather than learning the full high-fidelity stochastic solution operator directly, the proposed framework learns probabilistic residual operators that characterize the discrepancy between low- and high-fidelity solutions. By parameterizing the vector field in flow matching using neural operators conditioned on both the known system input and low-fidelity solution, the framework amortizes probabilistic inference across input conditions while enabling uncertainty-aware and resolution-generalizable predictions across spatial discretizations. Numerical experiments on stochastic advection, Burgers', and Darcy flow systems demonstrate that the residual-augmented formulation improves predictive accuracy under the same high-fidelity data budget, while the probabilistic operator learning formulation enables accurate characterization of uncertainty in low-data regimes compared to learning high-fidelity stochastic operators directly from data.

Markov Chain Monte Carlo (MCMC), Laplace approximation (LA) and variational inference (VI) methods are popular approaches to Bayesian inference, each with trade-offs between computational cost and accuracy. However, a theoretical understanding of these differences is missing, particularly when both the sample size $n$ and the dimension $d$ are large. LA and Gaussian VI are justified by Bernstein-von Mises (BvM) theorems, and recent work has derived the characteristic condition $n\gg d^2$ for their validity, improving over the condition $n\gg d^3$. In this paper, we show for linear, logistic and Poisson regression that for $n\gtrsim d$, MCMC attains the same complexity scaling in $n$, $d$ as first-order optimization algorithms, up to sub-polynomial factors. Thus MCMC is competitive with LA and Gaussian VI in complexity, under a scaling between $n$ and $d$ more general than BvM regimes. Our complexities apply to appropriately scaled priors that are not necessarily Gaussian-tailed, including Student-$t$ and flat priors, with log-posteriors that are not necessarily globally concave or gradient-Lipschitz.

We propose a unified framework that employs variational inference (VI) with (conditional) normalizing flows (NFs) to train both between-model and within-model proposals for reversible jump Markov chain Monte Carlo, enabling efficient trans-dimensional Bayesian inference. In contrast to the transport reversible jump (TRJ) of Davies et al. (2023), which optimizes forward KL divergence using pilot samples from the complex target distribution, our approach minimizes the reverse KL divergence, requiring only samples from a simple base distribution and largely reducing computational cost. Especially, we develop a novel trans-dimensional VI method with conditional NFs to fit the conditional transport proposal of Davies et al. (2023). We use RealNVP flows to learn the model-specific transport maps used for constructing proposals so that the calculation is parallelizable. Our framework also provides accurate estimates of marginal likelihoods, which may facilitate efficient model comparison and help design rejection-free proposals. Extensive numerical studies demonstrate that the TRJ method trained under our framework achieves faster mixing compared to existing baselines.

Machine learning (ML) methods are highly flexible, but their ability to approximate the true data-generating process is fundamentally constrained by finite samples. We characterize a universal lower bound, the Limits-to-Learning Gap (LLG), quantifying the unavoidable discrepancy between a model's empirical fit and the population benchmark. Recovering the true population $R^2$, therefore, requires correcting observed predictive performance by this bound. Using a broad set of variables, including excess returns, yields, credit spreads, and valuation ratios, we find that the implied LLGs are large. This indicates that standard ML approaches can substantially understate true predictability in financial data. We also derive LLG-based refinements to the classic Hansen and Jagannathan (1991) bounds, analyze implications for parameter learning in general-equilibrium settings, and show that the LLG provides a natural mechanism for generating excess volatility.

We propose a robust and scalable variational Bayes (VB) framework designed to effectively handle contamination and outliers in dataset. Our approach partitions the data into $m$ disjoint subsets and formulates a joint optimization problem based on robust aggregation principles. A key insight is that the full posterior distribution is equivalent to the minimizer of the mean Kullback-Leibler (KL) divergence from the $m$-powered local posterior distributions. To enhance robustness, we replace the mean KL divergence with a min-max median formulation. The min-max formulation not only ensures consistency between the KL minimizer and the Evidence Lower Bound (ELBO) maximizer but also facilitates the establishment of improved statistical rates for the mean of variational posterior. We observe a notable discrepancy in the $m$-powered marginal log likelihood function contingent on the presence of local latent variables. To address this, we treat these two scenarios separately to guarantee the consistency of the aggregated variational posterior. Specifically, when local latent variables are present, we introduce an aggregate-and-rescale strategy. Theoretically, we provide a non-asymptotic analysis of our proposed posterior, incorporating a refined analysis of Bernstein-von Mises (BvM) theorem to accommodate a diverging number of subsets $m$. Our findings indicate that the two-stage approach yields a smaller approximation error compared to directly aggregating the $m$-powered local posteriors. Furthermore, we establish a nearly optimal statistical rate for the mean of the proposed posterior, advancing existing theories related to min-max median estimators. The efficacy of our method is demonstrated through extensive simulation studies.

This paper develops a continuous functional framework for treatment effects propagating through geographic space and economic networks. We derive a master equation from three independent economic foundations – heterogeneous agent aggregation, market equilibrium, and cost minimization – establishing that the framework rests on fundamental principles rather than ad hoc specifications. The framework nests conventional econometric models – autoregressive specifications, spatial autoregressive models, and network treatment effect models – as special cases, providing a bridge between discrete and continuous methods. A key theoretical result shows that the spatial-network interaction coefficient equals the mutual information between geographic and network coordinates, providing a parameter-free measure of channel complementarity. The Feynman-Kac representation characterizes treatment effects as accumulated policy exposure along stochastic paths representing economic linkages, connecting the continuous framework to event study methodology. The no-spillover case emerges as a testable restriction, creating a one-sided risk profile where correct inference is maintained regardless of whether spillovers exist. Monte Carlo simulations confirm that conventional estimators exhibit 25-38% bias when spillovers are present, while our estimator maintains correct inference across all configurations including the no-spillover case.

Parkinson's Disease (PD) is a devastating neurodegenerative disorder that affects millions of people around the globe. Many researchers are continuously working to understand PD and develop treatments to improve the condition of PD patients, which affects their day-to-day lives. Since the last decades, the treatment, Deep Brain Stimulation (DBS) has given promising results for motor symptoms by improving the quality of daily living of PD patients. In the methodology of the present study, we have utilized sophisticated statistical approaches such as Nonparametric, Semi-parametric, and robust Parametric survival analysis to extract useful and important information about the long-term survival outcomes of the patients who underwent DBS for PD. Finally, we were able to conclude that the probabilistic behavior of the survival time of female patients is statistically different from that of male patients. Furthermore, we have identified that the probabilistic behavior of the survival times of Female patients is characterized by the 3-parameter Lognormal distribution, while that of Male patients is characterized by the 3-parameter Weibull distribution. More importantly, we have found that the Female patients have higher survival compared to the Male patients after conducting a robust parametric survival analysis. Using the semi-parametric COX-PH, we found that the initial implant of the right side leads to a high frequency of events occurring for the female patients with a bad prognostic factor, while for the male patients, a low events occurs with a good prognostic factor. Furthermore, we have found an interaction term between the number of revisions and the initial size of the implant, which increases the frequency of events occurring for the Male patients with a bad prognostic factor.

Modeling response surfaces with abrupt jumps and discontinuities remains a major challenge across scientific and engineering domains. Although Gaussian process models excel at capturing smooth nonlinear relationships, their stationarity assumptions limit their ability to adapt to sudden input-output variations. Existing nonstationary extensions, particularly those based on domain partitioning, often struggle with boundary inconsistencies, sensitivity to outliers, and scalability issues in higher-dimensional settings, leading to reduced predictive accuracy and unreliable parameter estimation. To address these challenges, this paper proposes the Robust Local Gaussian Process (RLGP) model, a framework that integrates adaptive nearest-neighbor selection with a sparsity-driven robustification mechanism. Unlike existing methods, RLGP leverages an optimization-based mean-shift adjustment after a multivariate perspective transformation combined with local neighborhood modeling to mitigate the influence of outliers. This approach improves predictive accuracy near discontinuities while enhancing robustness to data heterogeneity. Comprehensive evaluations on real-world datasets show that RLGP consistently delivers high predictive accuracy and maintains competitive computational efficiency, especially in scenarios with sharp transitions and complex response structures. Scalability tests further confirm RLGP's stability and reliability in higher-dimensional settings, where other methods struggle. These results establish RLGP as an effective and practical solution for modeling nonstationary and discontinuous response surfaces across a wide range of applications.

Data attribution aims to explain model predictions by estimating how they would change if certain training points were removed, and is used in a wide range of applications, from interpretability and credit assignment to unlearning and privacy. Even in the relatively simple case of logistic regressions, existing mathematical analyses of leading data attribution methods such as Influence Functions (IF) and single Newton Step (NS) remain limited in two key ways. First, they rely on global strong convexity assumptions which are often not satisfied in practice. Second, the resulting bounds scale very poorly with the number of parameters ($d$) and the number of samples removed ($k$). As a result, these analyses are not tight enough to answer fundamental questions such as "what is the asymptotic scaling of the errors of each method?" or "which of these methods is more accurate for a given dataset?" In this paper, we introduce a new analysis of the NS and IF data attribution methods for convex learning problems. To the best of our knowledge, this is the first analysis of these questions that does not assume global strong convexity and also the first explanation of [KATL19] and [RH25a]'s observation that NS data attribution is often more accurate than IF. We prove that for sufficiently well-behaved logistic regressions, our bounds are asymptotically tight up to poly-logarithmic factors, yielding scaling laws for the errors in the average-case sample removals. \[ \mathbb{E}_{T \subseteq [n],\, |T| = k} \bigl[ \|\hat{\theta}_T - \hat{\theta}_T^{\mathrm{NS}}\|_2 \bigr] = \widetilde{\Theta}\!\left(\frac{k d}{n^2}\right), \qquad \mathbb{E}_{T \subseteq [n],\, |T| = k} \bigl[ \|\hat{\theta}_T^{\mathrm{NS}} - \hat{\theta}_T^{\mathrm{IF}}\|_2 \bigr] = \widetilde{\Theta}\!\left( \frac{(k + d)\sqrt{k d}}{n^2} \right). \]

We resolve a 30-year-old open problem concerning the power of unlabeled data in online learning by tightly quantifying the gap between transductive and standard online learning. In the standard setting, the optimal mistake bound is characterized by the Littlestone dimension $d$ of the concept class $H$ (Littlestone 1987). We prove that in the transductive setting, the mistake bound is at least $\Omega(\sqrt{d})$. This constitutes an exponential improvement over previous lower bounds of $\Omega(\log\log d)$, $\Omega(\sqrt{\log d})$, and $\Omega(\log d)$, due respectively to Ben-David, Kushilevitz, and Mansour (1995, 1997) and Hanneke, Moran, and Shafer (2023). We also show that this lower bound is tight: for every $d$, there exists a class of Littlestone dimension $d$ with transductive mistake bound $O(\sqrt{d})$. Our upper bound also improves upon the best known upper bound of $(2/3)d$ from Ben-David, Kushilevitz, and Mansour (1997). These results establish a quadratic gap between transductive and standard online learning, thereby highlighting the benefit of advance access to the unlabeled instance sequence. This contrasts with the PAC setting, where transductive and standard learning exhibit similar sample complexities.

Distributionally robust optimization (DRO) has emerged as a powerful paradigm for reliable decision-making under uncertainty. This paper focuses on DRO with ambiguity sets defined via the Sinkhorn discrepancy: an entropy-regularized Wasserstein distance, referred to as Sinkhorn DRO. Existing work primarily addresses Sinkhorn DRO from a dual perspective, leveraging its formulation as a conditional stochastic optimization problem, for which many stochastic gradient methods are applicable. However, the theoretical analyses of such methods often rely on the boundedness of the loss function, and it is indirect to obtain the worst-case distribution associated with Sinkhorn DRO. In contrast, we study Sinkhorn DRO from the primal perspective, by reformulating it as a bilevel program with several infinite-dimensional lower-level subproblems over probability space. This formulation enables us to simultaneously obtain the optimal robust decision and the worst-case distribution, which is valuable in practical settings, such as generating stress-test scenarios or designing robust learning algorithms. We propose both double-loop and single-loop sampling-based algorithms with theoretical guarantees to solve this bilevel program. Finally, we demonstrate the effectiveness of our approach through a numerical study on adversarial classification.

Partial Differential Equations (PDEs) are central to modeling complex systems across physical, biological, and engineering domains, yet traditional numerical methods often struggle with high-dimensional or complex problems. Physics-Informed Neural Networks (PINNs) have emerged as an efficient alternative by embedding physics-based constraints into deep learning frameworks, but they face challenges in achieving high accuracy and handling complex boundary conditions. In this work, we extend the Time-Evolving Natural Gradient (TENG) framework to address Dirichlet boundary conditions, integrating natural gradient optimization with numerical time-stepping schemes, including Euler and Heun methods, to ensure both stability and accuracy. By incorporating boundary condition penalty terms into the loss function, the proposed approach enables precise enforcement of Dirichlet constraints. Experiments on the heat equation demonstrate the superior accuracy of the Heun method due to its second-order corrections and the computational efficiency of the Euler method for simpler scenarios. This work establishes a foundation for extending the framework to Neumann and mixed boundary conditions, as well as broader classes of PDEs, advancing the applicability of neural network-based solvers for real-world problems.

Medical data often exhibit characteristics that make cluster analysis particularly challenging, such as missing values, outliers, and cluster features like skewness. Typically, such data would need to be preprocessed – by cleaning outliers and missing values – before clustering could be performed. However, these preliminary steps rely on objective functions different from those used in the clustering stage. In this paper, we propose a unified model-based clustering approach that simultaneously handles atypical observations, missing values, and cluster-wise skewness within a single framework. Each cluster is modelled using a contaminated multivariate skew-normal distribution – a convenient two-component mixture of multivariate skew-normal densities – in which one component represents the main data (the "bulk") and the other captures potential outliers. From an inferential perspective, we implement and use a variant of the EM algorithm to obtain the maximum likelihood estimates of the model parameters. Simulation studies demonstrate that the proposed model outperforms existing approaches in both clustering accuracy and outlier detection, across low- and high-dimensional settings, even in the presence of substantial missingness. The method is further applied to the Cleveland Children's Sleep and Health Study (CCSHS), a dataset characterised by incomplete observations. Without any preprocessing, the proposed approach identifies five distinct groups of sleepers, revealing meaningful differences in sleeper typologies.

Classical statistical learning theory predicts a U-shaped relationship between test loss and model capacity, driven by the bias-variance trade-off. Recent advances in modern machine learning have revealed a more complex pattern, double-descent, in which test loss, after peaking near the interpolation threshold, decreases again as model capacity continues to grow. While this behavior has been extensively analyzed in regression and classification, its manifestation in survival analysis remains unexplored. This study investigates overparametrization in four representative survival models: DeepSurv, PC-Hazard, Nnet-Survival, and N-MTLR. We rigorously define interpolation and finite-norm interpolation, two key characteristics of loss-based models to understand double-descent. We then show the existence (or absence) of (finite-norm) interpolation of all four models. Our findings clarify how likelihood-based losses and model implementation jointly determine the feasibility of interpolation and show that overparametrization should not be regarded as benign for survival models. All theoretical results are supported by numerical experiments that highlight the distinct generalization behaviors of survival models.

When individuals engage in social or physical interactions, a unit's outcome may depend on the treatments received by others. In such interference environments, we provide a unified framework characterizing a broad class of spillover estimands as weighted averages of unit-to-unit spillover effects, with estimand-specific weights. We then develop design-based weighted least squares (WLS) estimators for both average and conditional spillover effects. We introduce three nonparametric estimators under the dyadic, sender, and receiver perspectives, which distribute the estimand weights differently across the outcome vector, design matrix, and weight matrix. For the average-type estimands, we show that all three estimators are equivalent to the Hajek estimator. For conditional spillover effects, we establish conditions under which the estimands are consistent for the target conditional spillover effects. We further derive concentration inequalities, a central limit theorem, and conservative variance estimators in an asymptotic regime where both the number of clusters and cluster sizes grow.