Papers

We propose a two-sample mean test based on the Bayes factor with non-informative priors, specifically designed for scenarios where the dimension $p$ grows with the sample size $n$ with a linear rate $p/n \to c_1 \in (0, \infty)$. We establish the asymptotic normality of the test statistic and the asymptotic power. Through extensive simulations, we demonstrate that the proposed test performs competitively against several existing methods, particularly when the marginal variances of the individual features are heterogeneous and when the sample size is small. Furthermore, our test remains robust under distribution misspecification. The proposed method not only effectively detects both sparse and non-sparse differences in mean vectors but also maintains a well-controlled type I error rate, even in small-sample scenarios. We also demonstrate the performance of our proposed test using the small round blue cell tumors (SRBCT) dataset.

Recently, varextropy has been introduced as a new dispersion index and a measure of information. In this article, we derive the generating function of extropy and present its infinite series representation. Furthermore, we propose new variability measures: the inaccuracy and weighted inaccuracy measures between two random variables based on varextropy and we investigate their properties. We also obtain lower bounds for the inaccuracy measure and compare them with each other. In addition, we introduce a discrimination measure based on varextropy and employ it both for comparing probability distributions and for assessing the goodness of fit of distributions to data model selection problems. The proposed measure is also compared with the dispersion index derived from the Kullback-Leibler divergence given in Balakrishnan et al. (2024).

We observe an unknown function of $d$ variables $f(\boldsymbol{t})$, $\boldsymbol{t} \in[0,1]^d$, in the Gaussian white noise model of intensity $\varepsilon>0$. We assume that the function $f$ is regular and that it is a sum of $k$-variate functions, where $k$ varies from $1$ to $s$ ($1\leq s\leq d$). These functions are unknown to us and only a few of them are nonzero. In this article, we address the problem of identifying the nonzero function components of $f$ almost fully in the case when $d=d_\varepsilon\to \infty$ as $\varepsilon\to 0$ and $s$ is either fixed or $s=s_\varepsilon\to \infty$, $s=o(d)$ as $\varepsilon\to 0$. This may be viewed as a variable selection problem. We derive the conditions when almost full variable selection in the model at hand is possible and provide a selection procedure that achieves this type of selection. The procedure is adaptive to the level of sparsity described by the sparsity index $\beta\in(0,1)$. We also derive conditions that make almost full variable selection in the model of our interest impossible. In view of these conditions, the proposed selector is seen to perform asymptotically optimal. The theoretical findings are illustrated numerically.

Inference for locally stationary time series is challenging because the associated hypotheses form an uncountable collection over a continuous time interval, making pointwise false discovery rate (FDR) control inadequate for simultaneous statistical guarantees. We introduce a novel asymptotically uniform false discovery rate (AuFDR), defined as the expectation of the $L_r$-norm of the false discovery proportion (FDP) process where $r$ is allowed to diverge, to quantify and control false discoveries uniformly over time. To operationalize AuFDR control, we develop an inferential framework for time-varying correlations in high-dimensional nonstationary time series that allows for non-Gaussianity, nonlinearity and possible jumps in mean functions. The proposed approach combines robust difference-based estimators with a multiplier-bootstrap procedure to construct uniformly valid time-varying $P$-values. Based on these $P$-values, we propose a time-varying Benjamini–Yekutieli procedure for controlling the AuFDR under arbitrary dependence and establish its asymptotic validity. Extensive simulations demonstrate the finite-sample performance of the proposed method in controlling the AuFDR. Applications to EEG data and financial time-series data illustrate its practical utility.

Forward Osmosis (FO) is a promising low-energy membrane separation technology, but challenges in accurately modelling its water flux (Jw) persist due to complex internal mass transfer phenomena. Traditional mechanistic models struggle with empirical parameter variability, while purely data-driven models lack physical consistency and rigorous uncertainty quantification (UQ). This study introduces a novel Robust Hybrid Physics-ML framework employing Gaussian Process Regression (GPR) for highly accurate, uncertainty-aware Jw prediction. The core innovation lies in training the GPR on the residual error between the detailed, non-linear FO physical model prediction (Jw_physical) and the experimental water flux (Jw_actual). Crucially, we implement a full UQ methodology by decomposing the total predictive variance (sigma2_total) into model uncertainty (epistemic, from GPR's posterior variance) and input uncertainty (aleatoric, analytically propagated via the Delta method for multi-variate correlated inputs). Leveraging the inherent strength of GPR in low-data regimes, the model, trained on a meagre 120 data points, achieved a state-of-the-art Mean Absolute Percentage Error (MAPE) of 0.26% and an R2 of 0.999 on the independent test data, validating a truly robust and reliable surrogate model for advanced FO process optimization and digital twin development.

Many scientific areas, from computer science to the environmental sciences and finance, give rise to multivariate time series which exhibit long memory, or loosely put, a slow decay in their autocorrelation structure. Efficient modelling and estimation in such settings is key for a number of analysis tasks, such as accurate prediction. However, traditional approaches for modelling such data, for example long memory vector autoregressive processes, are challenging even in modest dimensions, as the number of parameters grows quadratically with the number of modelled variables. Additionally, in many practical data settings, the observed series is accompanied by a (possibly inferred) network that provides information about the presence or absence of between-component associations via the graph edge topology. This article proposes two new models for capturing the dynamics of long memory time series where a network is accounted for. Our approach not only facilitates the analysis of graph-structured long memory time series, but also improves computational efficiency over traditional multivariate long memory models by leveraging the inherent low-dimensional parameter space by adapting likelihood-based estimation algorithms to the network setting. Simulation studies show that our proposed estimation is more stable than traditional models, and is able to tackle data scenarios where current models fail due to computational challenges. While widely applicable, here we demonstrate the efficacy of our proposed models on datasets arising in environmental science and finance.

We consider a regression setting where observations are collected in different environments modeled by different data distributions. The field of out-of-distribution (OOD) generalization aims to design methods that generalize better to test environments whose distributions differ from those observed during training. One line of such works has proposed to minimize the maximum risk across environments, a principle that we refer to as MaxRM (Maximum Risk Minimization). In this work, we introduce variants of random forests based on the principle of MaxRM. We provide computationally efficient algorithms and prove statistical consistency for our primary method. Our proposed method can be used with each of the following three risks: the mean squared error, the negative reward, and the regret (which quantifies the excess risk relative to the best predictor). For MaxRM with regret as the risk, we prove a novel out-of-sample guarantee over unseen test distributions. Finally, we evaluate the proposed methods on both simulated and real-world data.

This paper introduces a new probabilistic framework for supervised learning in neural systems. It is designed to model complex, uncertain systems whose random outputs are strongly non-Gaussian given deterministic inputs. The architecture itself is a random object stochastically generated by a latent anisotropic Gaussian random field defined on a compact, boundaryless, multiply-connected manifold. The goal is to establish a novel conceptual and mathematical framework in which neural architectures are realizations of a geometry-aware, field-driven generative process. Both the neural topology and synaptic weights emerge jointly from a latent random field. A reduced-order parameterization governs the spatial intensity of an inhomogeneous Poisson process on the manifold, from which neuron locations are sampled. Input and output neurons are identified via extremal evaluations of the latent field, while connectivity is established through geodesic proximity and local field affinity. Synaptic weights are conditionally sampled from the field realization, inducing stochastic output responses even for deterministic inputs. To ensure scalability, the architecture is sparsified via percentile-based diffusion masking, yielding geometry-aware sparse connectivity without ad hoc structural assumptions. Supervised learning is formulated as inference on the generative hyperparameters of the latent field, using a negative log-likelihood loss estimated through Monte Carlo sampling from single-observation-per-input datasets. The paper initiates a mathematical analysis of the model, establishing foundational properties such as well-posedness, measurability, and a preliminary analysis of the expressive variability of the induced stochastic mappings, which support its internal coherence and lay the groundwork for a broader theory of geometry-driven stochastic learning.

This paper is concerned with differentiable resampling in the context of sequential Monte Carlo (e.g., particle filtering). Drawing on reparametrisation, we propose a new resampling method that is informative and instantly differentiable, based on a training-free diffusion model surrogate. We theoretically prove that our diffusion resampling method provides a consistent resampling distribution, and we show empirically that it outperforms the state-of-the-art differentiable resampling methods on multiple filtering and parameter estimation benchmarks. Finally, we show that it achieves competitive end-to-end performance when used in learning a complex dynamics-decoder model with high-dimensional image observations.

This study introduces a new family of probability distributions, termed the alpha power Harris-generalized (APHG) family. The generator arises by incorporating two shape parameters from the Harris-G framework into the alpha power transformation, resulting in a more flexible class for modelling survival and reliability data. A special member of this family, obtained using the two-parameter Burr XII distribution as the baseline, is developed and examined in detail. Several analytical properties of the proposed alpha power Harris Burr XII (APHBXII) model are derived, which include closed-form expressions for its moments, mean and median deviations, Bonferroni and Lorenz curves, order statistics, and Renyi and Tsallis entropies. Parameter estimation is performed via maximum likelihood, and a Monte Carlo simulation study is carried out to assess the finite-sample performance of the estimators. In addition, three real lifetime datasets are analyzed to evaluate the empirical performance of the APHBXII distribution relative to four competing models. The results show that the five-parameter APHBXII model provides superior fit across all datasets, as supported by model-selection criteria and goodness-of-fit statistics.

We analyze prediction error in stochastic dynamical systems with memory, focusing on generalized Langevin equations (GLEs) formulated as stochastic Volterra equations. We establish that, under a strongly convex potential, trajectory discrepancies decay at a rate determined by the decay of the memory kernel and are quantitatively bounded by the estimation error of the kernel in a weighted norm. Our analysis integrates synchronized noise coupling with a Volterra comparison theorem, encompassing both subexponential and exponential kernel classes. For first-order models, we derive moment and perturbation bounds using resolvent estimates in weighted spaces. For second-order models with confining potentials, we prove contraction and stability under kernel perturbations using a hypocoercive Lyapunov-type distance. This framework accommodates non-translation-invariant kernels and white-noise forcing, explicitly linking improved kernel estimation to enhanced trajectory prediction. Numerical examples validate these theoretical findings.

We propose a Bayesian framework for multilayer song similarity networks and apply it to the complete studio discographies of the "Big 4" of thrash metal (Metallica, Slayer, Megadeth, Anthrax). Starting from raw audio, we construct four feature-specific layers (loudness, brightness, tonality, rhythm), augment them with song exogenous information, and represent each layer as a k-nearest neighbor graph. We then fit a family of hierarchical probit models with global and layer-specific baselines, node- and layer-specific sociability effects, dyadic covariates, and alternative forms of latent structure (bilinear, distance-based, and stochastic block communities), comparing increasingly flexible specifications using posterior predictive checks, discrimination and calibration metrics (AUC, Brier score, log-loss), and information criteria (DIC, WAIC). Across all bands, the richest stochastic block specification attains the best predictive performance and posterior predictive fit, while revealing sparse but structured connectivity, interpretable covariate effects (notably album membership and temporal proximity), and latent communities and hubs that cut across albums and eras. Taken together, these results illustrate how Bayesian multilayer network models can help organize high-dimensional audio and text features into coherent, musically meaningful patterns.

Drift diffusion models (DDMs) have found widespread use in computational neuroscience, cognitive science, mathematical psychology as well as other fields. They model evidence accumulation in simple decision tasks as a stochastic process drifting towards decision barriers. In models where the drift is both time-varying within a trial and variable across trials, the high computational cost for accurate likelihood evaluation has often led to the use of a computationally convenient surrogate for parameter inference, the time-averaged drift approximation (TADA). In each trial, TADA assumes that the time-varying drift rate can be replaced by its temporal average throughout the trial. This approach enables fast parameter inference using analytical likelihood formulas for DDMs with constant drift. In this work, we show that such an estimator is inconsistent: it does not converge to the true drift, posing a risk of biasing scientific conclusions when parameter estimates are obtained by TADA and similar approximations. We provide an elementary proof of this inconsistency in what is perhaps the simplest possible setting: a Brownian motion with piecewise constant drift hitting a one-sided upper boundary. Furthermore, numerical examples based on an attentional DDM (aDDM) show that using TADA leads to systematic misestimation of attentional effects in decision making and can lead to false conclusions in scientific hypothesis testing.

The property of learning-curve monotonicity, highlighted in a recent series of work by Loog, Mey and Viering, describes algorithms which only improve in average performance given more data, for any underlying data distribution within a given family. We establish the first nontrivial monotonicity guarantees for the maximum likelihood estimator in a variety of well-specified parametric settings. For sequential prediction with log loss, we show monotonicity (in fact complete monotonicity) of the forward KL divergence for Gaussian vectors with unknown covariance and either known or unknown mean, as well as for Gamma variables with unknown scale parameter. The Gaussian setting was explicitly highlighted as open in the aforementioned works, even in dimension 1. Finally we observe that for reverse KL divergence, a folklore trick yields monotonicity for very general exponential families. All results in this paper were derived by variants of GPT-5.2 Pro. Humans did not provide any proof strategies or intermediate arguments, but only prompted the model to continue developing additional results, and verified and transcribed its proofs.

Detection limits are common in biomedical and environmental studies, where key covariates or outcomes are censored below an assay-specific threshold. Standard approaches such as complete-case analysis, single-value substitution, and parametric Tobit-type models are either inefficient or sensitive to distributional misspecification. We study semiparametric rank-based regression models as robust alternatives to parametric mean-based counterparts for censored responses under detection limits. Our focus is on accelerated failure time (AFT) type formulations, where rank-based estimating equations yield consistent slope estimates without specifying the error distribution. We develop a unifying simulation framework that generates left- and right-censored data under several data-generating mechanisms, including normal, Weibull, and log-normal error structures, with detection limits or administrative censoring calibrated to target censoring rates between 10% and 60%. Across scenarios, we compare semiparametric AFT estimators with parametric Weibull AFT, Tobit, and Cox proportional hazards models in terms of bias, empirical variability, and relative efficiency. Numerical results show that parametric models perform well only under correct specification, whereas rank-based semiparametric AFT estimators maintain near-unbiased covariate effects and stable precision even under heavy censoring and distributional misspecification. These findings support semiparametric rank-based regression as a practical default for censored regression with detection limits when the error distribution is uncertain. Keywords: Semiparametric models, Estimating equations, Left censoring, Right censoring, Tobit regression, Efficiency

We analyze gradient descent with randomly weighted data points in a linear regression model, under a generic weighting distribution. This includes various forms of stochastic gradient descent, importance sampling, but also extends to weighting distributions with arbitrary continuous values, thereby providing a unified framework to analyze the impact of various kinds of noise on the training trajectory. We characterize the implicit regularization induced through the random weighting, connect it with weighted linear regression, and derive non-asymptotic bounds for convergence in first and second moments. Leveraging geometric moment contraction, we also investigate the stationary distribution induced by the added noise. Based on these results, we discuss how specific choices of weighting distribution influence both the underlying optimization problem and statistical properties of the resulting estimator, as well as some examples for which weightings that lead to fast convergence cause bad statistical performance.

Topology identification and inference of processes evolving over graphs arise in timely applications involving brain, transportation, financial, power, as well as social and information networks. This chapter provides an overview of graph topology identification and statistical inference methods for multidimensional relational data. Approaches for undirected links connecting graph nodes are outlined, going all the way from correlation metrics to covariance selection, and revealing ties with smooth signal priors. To account for directional (possibly causal) relations among nodal variables and address the limitations of linear time-invariant models in handling dynamic as well as nonlinear dependencies, a principled framework is surveyed to capture these complexities through judiciously selected kernels from a prescribed dictionary. Generalizations are also described via structural equations and vector autoregressions that can exploit attributes such as low rank, sparsity, acyclicity, and smoothness to model dynamic processes over possibly time-evolving topologies. It is argued that this approach supports both batch and online learning algorithms with convergence rate guarantees, is amenable to tensor (that is, multi-way array) formulations as well as decompositions that are well-suited for multidimensional network data, and can seamlessly leverage high-order statistical information.

Lung cancer (LC) is a leading cause of cancer-related mortality in the United States. Accurate prediction of LC mortality rates is crucial for guiding targeted interventions and addressing health disparities. Although traditional regression-based models have been commonly used, explainable machine learning models may offer enhanced predictive accuracy and deeper insights into the factors influencing LC mortality. This study applied three models: random forest (RF), gradient boosting regression (GBR), and linear regression (LR) to predict county-level LC mortality rates across the United States. Model performance was evaluated using R-squared and root mean squared error (RMSE). Shapley Additive Explanations (SHAP) values were used to determine variable importance and their directional impact. Geographic disparities in LC mortality were analyzed through Getis-Ord (Gi*) hotspot analysis. The RF model outperformed both GBR and LR, achieving an R2 value of 41.9% and an RMSE of 12.8. SHAP analysis identified smoking rate as the most important predictor, followed by median home value and the percentage of the Hispanic ethnic population. Spatial analysis revealed significant clusters of elevated LC mortality in the mid-eastern counties of the United States. The RF model demonstrated superior predictive performance for LC mortality rates, emphasizing the critical roles of smoking prevalence, housing values, and the percentage of Hispanic ethnic population. These findings offer valuable actionable insights for designing targeted interventions, promoting screening, and addressing health disparities in regions most affected by LC in the United States.

We introduce the cumulative residual Mathai–Haubold entropy (CRMHE) and investigate its properties. We then propose a dynamic counterpart, the dynamic cumulative residual Mathai–Haubold entropy (DCRMHE), and establish its uniqueness in characterizing the distribution function. Non-parametric estimators for the CRMHE and DCRMHE are developed based on the kernel density estimation of the survival function. The efficacy of the estimators is assessed through a comprehensive Monte Carlo simulation study. The relevance of the proposed DCRMHE estimator is illustrated using two real-world datasets: on the failure times of 70 aircraft windshields and failure times of 40 randomly selected mechanical switches.

We present an approach to denoising spatial transcriptomics images that is particularly effective for uncovering cell identities in the regime of ultra-low sequencing depths, and also allows for interpolation of gene expression. The method – Spatial Transcriptomics via Adaptive Regularization and Kernels (STARK) – augments kernel ridge regression with an incrementally adaptive graph Laplacian regularizer. In each iteration, we (1) perform kernel ridge regression with a fixed graph to update the image, and (2) update the graph based on the new image. The kernel ridge regression step involves reducing the infinite dimensional problem on a space of images to finite dimensions via a modified representer theorem. Starting with a purely spatial graph, and updating it as we improve our image makes the graph more robust to noise in low sequencing depth regimes. We show that the aforementioned approach optimizes a block-convex objective through an alternating minimization scheme wherein the sub-problems have closed form expressions that are easily computed. This perspective allows us to prove convergence of the iterates to a stationary point of this non-convex objective. Statistically, such stationary points converge to the ground truth with rate $\mathcal{O}(R^{-1/2})$ where $R$ is the number of reads. In numerical experiments on real spatial transcriptomics data, the denoising performance of STARK, evaluated in terms of label transfer accuracy, shows consistent improvement over the competing methods tested.

The advantages of adaptive experiments have led to their rapid adoption in economics, other fields, and among practitioners. However, adaptive experiments pose challenges for causal inference. This note suggests a BOLS (batched ordinary least squares) test statistic for inference on treatment effects in adaptive experiments. The statistic provides a precision-equalizing aggregation of per-period treatment-control differences under heteroskedasticity. The combined test statistic is a linear combination of heteroskedastic per-period z-statistics and can be used to construct asymptotically valid confidence intervals. We provide simulation results comparing rejection rates in the typical case with few treatment periods and few (or many) observations per batch.

Reliable data-driven estimation of Shannon entropy from small data sets, where the number of examples is potentially smaller than the number of possible outcomes, is a critical matter in several applications. In this paper, we introduce a discrete entropy estimator, where we use the decomposability property in combination with estimations of the missing mass and the number of unseen outcomes to compensate for the negative bias induced by them. Experimental results show that the proposed method outperforms some classical estimators in undersampled regimes, and performs comparably with some well-established state-of-the-art estimators.

Learning Analytics (LA) has rapidly expanded through practical and technological innovation, yet its foundational identity has remained theoretically under-specified. This paper addresses this gap by proposing the first axiomatic theory that formally defines the essential structure, scope, and limitations of LA. Derived from the psychological definition of learning and the methodological requirements of LA, the framework consists of five axioms specifying discrete observation, experience construction, state transition, and inference. From these axioms, we derive a set of theorems and propositions that clarify the epistemological stance of LA, including the inherent unobservability of learner states, the irreducibility of temporal order, constraints on reachable states, and the impossibility of deterministically predicting future learning. We further define LA structure and LA practice as formal objects, demonstrating the sufficiency and necessity of the axioms and showing that diverse LA approaches – such as Bayesian Knowledge Tracing and dashboards – can be uniformly explained within this framework. The theory provides guiding principles for designing analytic methods and interpreting learning data while avoiding naive behaviorism and category errors by establishing an explicit theoretical inference layer between observations and states. This work positions LA as a rigorous science of state transition systems based on observability, establishing the theoretical foundation necessary for the field's maturation as a scholarly discipline.

We establish a universal framework for concentration inequalities based on invariance under diffeomorphism groups. Given a probability measure $\mu$ on a space $E$ and a diffeomorphism $\psi: E \to F$, concentration properties transfer covariantly: if the pushforward $\psi_*\mu$ concentrates, so does $\mu$ in the pullback geometry. This reveals that classical concentration inequalities – Hoeffding, Bernstein, Talagrand, Gaussian isoperimetry – are manifestations of a single principle of geometric invariance. The choice of coordinate system $\psi$ becomes a free parameter that can be optimized. We prove that for any distribution class $\Pc$, there exists an optimal diffeomorphism $\psi^*$ minimizing the concentration constant, and we characterize $\psi^*$ in terms of the Fisher-Rao geometry of $\Pc$. We establish strict improvement theorems: for heavy-tailed or multiplicative data, the optimal $\psi$ yields exponentially tighter bounds than the identity. We develop the full theory including transportation-cost inequalities, isoperimetric profiles, and functional inequalities, all parametrized by the diffeomorphism group $\Diff(E)$. Connections to information geometry (Amari's $\alpha$-connections), optimal transport with general costs, and Riemannian concentration are established. Applications to robust statistics, multiplicative models, and high-dimensional inference demonstrate that coordinate optimization can improve statistical efficiency by orders of magnitude.

Dynamic Treatment Regimes (DTRs) provide a systematic framework for optimizing sequential decision-making in chronic disease management, where therapies must adapt to patients' evolving clinical profiles. Inverse probability weighting (IPW) is a cornerstone methodology for estimating regime values from observational data due to its intuitive formulation and established theoretical properties, yet standard IPW estimators face significant limitations, including variance instability and data inefficiency. A fundamental but underexplored source of inefficiency lies in the strict alignment requirement between observed and target treatment trajectories, which fails to account for partial compatibility and discards substantial information from individuals with only minimal deviations from the regime. We propose two novel methodologies that relax the strict inclusion rule through flexible compatibility mechanisms. Both methods provide computationally tractable alternatives that can be easily integrated into existing IPW workflows, offering more efficient approaches to DTR estimation. Theoretical analysis demonstrates that both estimators preserve consistency while achieving superior finite-sample efficiency compared to standard IPW, and comprehensive simulation studies confirm improved stability. We illustrate the practical utility of our methods through an application to HIV treatment data from the AIDS Clinical Trials Group Study 175 (ACTG175).

We develop TwinKernel methods for nonparametric estimation of intensity functions of point processes. Building on the general TwinKernel framework and combining it with martingale techniques for counting processes, we construct estimators that adapt to orbital regularity of the intensity function. Given a point process $N$ with intensity $\lambda$ and a cyclic group $G = \langle\varphi\rangle$ acting on the time/space domain, we transport kernels along group orbits to create a hierarchy of smoothed Nelson-Aalen type estimators. Our main results establish: (i) uniform consistency via martingale concentration inequalities; (ii) optimal convergence rates for intensities in twin-Hölder classes, with rates depending on the effective dimension $d_{\mathrm{eff}}$; (iii) adaptation to unknown smoothness through penalized model selection; (iv) automatic boundary bias correction via local polynomial extensions in twin coordinates; (v) minimax lower bounds showing rate optimality. We apply the methodology to hazard rate estimation under random censoring, where periodicity or other orbital structure in the hazard may arise from circadian rhythms, seasonal effects, or treatment schedules. Martingale central limit theorems yield asymptotic confidence bands. Simulation studies demonstrate 3–7$\times$ improvements over classical kernel hazard estimators when the intensity exhibits orbital regularity.

The remarkable success of AlphaFold2 in providing accurate atomic-level prediction of protein structures from their amino acid sequence has transformed approaches to the protein folding problem. However, its core paradigm of mapping one sequence to one structure may only be appropriate for single-fold proteins with one stable conformation. Metamorphic proteins, which can adopt multiple distinct conformations, have conformational diversity that cannot be adequately modeled by AlphaFold2. Hence, classifying whether a given protein is metamorphic or single-fold remains a critical challenge for both laboratory experiments and computational methods. To address this challenge, we developed a novel classification framework by re-purposing AlphaFold2 to generate conformational ensembles via a multiple sequence alignment sampling method. From these ensembles, we extract a comprehensive set of features characterizing the conformational ensemble's modality and structural dispersion. A random forest classifier trained on a carefully curated benchmark dataset of known metamorphic and single-fold proteins achieves a mean AUC of 0.869 with cross-validation, demonstrating the effectiveness of our integrated approach. Furthermore, by applying our classifier to 600 randomly sampled proteins from the Protein Data Bank, we identified several potential metamorphic protein candidates – including the 40S ribosomal protein S30, whose conformational change is crucial for its secondary function in antimicrobial defense. By combining AI-driven protein structure prediction with statistical learning, our work provides a powerful new approach for discovering metamorphic proteins and deepens our understanding of their role in their molecular function.

Physics-based battery modelling has emerged to accelerate battery materials discovery and performance assessment. Its success, however, is still hindered by difficulties in aligning models to experimental data. Bayesian approaches are a valuable tool to overcome these challenges, since they enable prior assumptions and observations to be combined in a principled manner that improves numerical conditioning. Here we introduce two new algorithms to the battery community, SOBER and BASQ, that greatly speed up Bayesian inference for parameterisation and model comparison. We showcase how Bayesian model selection allows us to tackle data observability, model identifiability, and data-informed model development together. We propose this approach for the search for battery models of novel materials.

Binary classification is one of the oldest, most prevalent, and studied problems in machine learning. However, the metrics used to evaluate model performance have received comparatively little attention. The area under the receiver operating characteristic curve (AUROC) has long been a standard choice for model comparison. Despite its advantages, AUROC is not always ideal, particularly for problems that are invariant to local exchange of classes (LxC), a new form of metric invariance introduced in this work. To address this limitation, we propose LxCIM (LxC-invariant metric), which is not only rank-based and invariant under local exchange of classes, but also intuitive, logically consistent, and always computable, while enabling more detailed analysis through the cumulative accuracy-decision rate curve. Moreover, LxCIM exhibits clear theoretical connections to AUROC, accuracy, and the area under the accuracy-decision rate curve (AUDRC). These relationships allow for multiple complementary interpretations: as a symmetric form of AUROC, a rank-based analogue of accuracy, or a more representative and more interpretable variant of AUDRC. Finally, we demonstrate the direct applicability of LxCIM to the bivariate causal discovery problem (which exhibits invariance to local exchange of classes) and show how it addresses the acknowledged limitations of existing metrics used in this field. All code and implementation details are publicly available at github.com/tiagobrogueira/Causal-Discovery-In-Exchangeable-Data.

We develop a unified framework for nonparametric functional estimation based on kernel transport along orbits of discrete group actions, which we term Twin Spaces. Given a base kernel $K$ and a group $G = \langle\varphi\rangle$ acting isometrically on the input space $E$, we construct a hierarchy of transported kernels $\{K_j\}_{j\geq 0}$ and a penalized model selection scheme satisfying a Kraft inequality. Our main contributions are threefold: (i) we establish non-asymptotic oracle inequalities for the penalized twin-kernel estimator with explicit constants; (ii) we introduce novel twin-regularity classes that capture smoothness along group orbits and prove that our estimator adapts to these classes; (iii) we show that the framework recovers classical minimax-optimal rates in the Euclidean setting while enabling improved rates when the target function exhibits orbital structure. The effective dimension $d_{\mathrm{eff}}$ governing the rates is characterized in terms of the quotient $G/L$, where $L$ is the subgroup preserving the base operation. Connections to wavelet methods, geometric quantization, and adaptive computation are discussed.