Papers

Early-phase oncology evidence is often summarized by asking whether a treatment is promising, but an encouraging signal can support several different next studies. We propose Evidence-to-Decision (E2D), a framework that organizes evidence around specific development actions. For each candidate action, E2D defines the contemplated future study, what would count as success, what evidence is needed before taking that step, what evidence is available now, and what assumptions connect the current and future settings. Statistical methods can then quantify action-specific support and sensitivity to those assumptions, while the final development choice remains multidisciplinary. We illustrate E2D in a rare pediatric oncology setting using predictive probabilities for same-setting single-arm expansion, same-setting randomized Phase II, movement to an earlier-line or frontline randomized study, and a separate de-escalation example. Calibrated support rules produced clinically distinct evidence states: intermediate evidence could support same-setting development without supporting frontline movement, safety concerns could narrow support despite strong activity, and stronger evidence could support several paths simultaneously. E2D therefore reframes the question from whether a treatment is generally promising to which next studies are supported by the evidence available now, what remains unsupported, and what additional evidence could change that assessment.

In platform trials, model-based approaches typically estimate treatment effects conditional on calendar time. However, scientific and regulatory interest often lies in treatment effects defined for a target population spanning multiple enrollment periods, requiring explicit consideration of how effects should be averaged across time. This raises two fundamental challenges. The first is the definition of the appropriate target estimand when combining data across multiple periods. The second is the selection of the estimator to be used. In this work, we examine conditional and marginal estimands in platform trials with time trends, and describe target populations of interest. To address the second challenge, we evaluate model-based, G-computation and augmented inverse probability weighting estimators, comparing their bias and variance. We discuss how the choice of estimand, target population and trial data used for estimation affects estimator performance.

We consider simulation-based Bayesian inference (SBI) for the parameters of models with intractable likelihoods but tractable forward simulation. Building on Gaussian mixtures-of-experts surrogates, as a first contribution we develop a sequential procedure to estimate the posterior, incorporating an importance sampling step that introduces a locally amortized likelihood-to-evidence ratio estimator. The local amortization allows the ratio estimator to be evaluated at new parameter–data pairs from a local (data-dependent) design, without re-estimation, which is particularly convenient for our second contribution. The second contribution is localized simulation-based calibration (SBC). Localized SBC provides a graphical diagnostic of calibration over neighborhoods that are deliberately broader than the available posterior, at low additional computational cost. Rather than repeatedly running the full inference procedure for each simulated pseudo-observation, localized SBC fits the local surrogate and ratio estimator only once for each local design, and reuses them across all pseudo-observations. This approach is particularly appealing for SBI applications. We evaluate both contributions on three case studies, including a real-data epidemiological application.

We study large deviations for rare clusters in sparse systems of moving particles. In the regime \(nr_n^d\to0\) and \(\rho_{k,n}=n^kr_n^{d(k-1)}\to\infty\), we prove a large deviation principle for the empirical measure of isolated \(k\)-particle trajectory clusters. The speed is \(\rho_{k,n}\), and the rate function is the relative entropy \(h(\,\cdot\,\mid\tau_k)\), where the finite reference measure \(\tau_k\) explicitly incorporates the underlying path law. As consequences, we derive free-energy variational formulas for bounded interactions and a hard-core constraint and identify the corresponding minimizing cluster law. The normalized optimizer provides the basis for Metropolis–Hastings sampling of interacting trajectory clusters. As an application motivated by chain formation and swarming in active particle systems, we calibrate the resulting stochastic model to experimental trajectories of magnetic micromotors and find that the fitted velocity scale varies systematically with particle size and magnetic forcing.

Days Alive and Out of Hospital (DAOH) and Days Alive at Home (DAH) are increasingly used as patient-centred primary outcomes in perioperative trials, and a recent editorial has advocated DAOH for its simplicity and reliance on routinely collected data. We argue that the choice between these endpoints affects power, treatment effect estimation and missing data, and should be made at the design stage with these consequences in mind. By treating days in nursing homes or rehabilitation facilities as equivalent to days at home, DAOH assumes that being out of hospital is a valid surrogate for good recovery, an assumption that patient perspectives, including those from the NOTACS trial, call into question. Choosing DAH over DAOH therefore involves a trade-off between measuring what matters to patients more accurately and the added burden of tracking discharge destination, including a greater risk of missing data. If an intervention affects only length of stay, readmission or mortality, the two endpoints yield the same expected treatment effect, and the extra burden of tracking discharge destination brings no benefit. If it instead enables more patients to return directly home rather than to a care facility, DAOH may conceal this benefit and lose statistical power. We propose a baseline-adjusted DAH, which counts only days spent in a setting representing an escalation of care relative to the patient's baseline, as a better proxy for recovery that also accommodates hospital-at-home and virtual ward services.

Generative models are transforming Bayesian computational imaging, yet the field still lacks physics-aware foundation models. Current practice falls into two camps. Large foundation image models are deployed as plug-and-play priors with zero-shot approximate likelihood guidance, which introduces significant bias and computational cost. Physics-aware generative models avoid this bias, but each is tied to a specific dataset, task and instrument. We introduce BAM (Bayesian Anything Model), a lightweight foundation model for few-step, physics-aware posterior sampling that generalises robustly to unseen data and tasks, zero-shot or with minimal finetuning. BAM upgrades the operator-conditioned Reconstruct Anything Model (RAM) backbone (Terris et al.) into a conditional flow map, so instrument physics is specified at inference time rather than fixed during training. BAM has just 36M parameters and is pre-trained jointly on large image corpora and libraries of forward operators. A single network then draws posterior samples in a few steps, with no likelihood approximation and no guidance weights to tune. Across linear inverse problems on FFHQ, AFHQ, LSUN, DIV2K and the Kohler camera-shake benchmark, BAM outperforms in just 3 steps both specialised models and leading zero-shot methods in sample quality, at a fraction of their computational cost. BAM gives the community an accessible entry point to generative computational imaging, lowers the economic and environmental cost of training imaging models, and opens a new path for research on physics-aware Bayesian computational imaging. Official page: https://bayesian-anything-model.github.io/

For estimating the mean vector of a $p$-variate normal distribution ($p \ge 3$) under quadratic loss, the positive-part James–Stein estimator dominates the original James–Stein estimator, but it possesses a non-smooth thresholding boundary. In this paper, by truncating the infinite series representation of the positive-part function to a finite sum of degree $m$, we propose a new class of smooth shrinkage estimators. We prove that the proposed estimator dominates the James–Stein estimator for any dimension $p \ge 3$, provided the truncation degree satisfies $m \ge 0.95\sqrt{p-2}$.

We study the problem of estimating the guidance that steers the distribution learned by a diffusion generative model toward a tilted target $q_0 \propto w\,p_0$ at inference time. Relying on the stochastic optimal control approach, we observe that the exact drift correction is the gradient of the logarithm of Doob's $h$-function, and we study the problem of estimating it from a sample. In the present paper, we assume that the score of the pretrained model is available, that the tilting weight is bounded and positive, and that the reference distribution has a bounded support, no smoothness of the weight is required. Introducing a penalized least-squares risk in which the penalty is the residual of the space-time harmonicity equation satisfied by the $h$-function, measured in a dual Sobolev norm, we derive high-probability bounds on the squared error of the resulting guidance estimate. Since the penalty vanishes at the target, the estimator is free of regularization bias, and in favourable scenarios its rate of convergence is faster than the minimax rate of estimating first-order derivatives of a smooth regression function. Assuming that $w$ is bounded and positive with $\mathbb{E}_{p_0}[w^{-\mathrm{s}}] < \infty$ for some $\mathrm{s} \in (0,\infty]$, and that the reference data are compactly supported, we prove that the guidance is estimable in squared $L^2$ at rate $\varepsilon_n^{\mathrm{s}/(\mathrm{s}+4)}$, where $\varepsilon_n = n^{-2(\beta-1)/(2(\beta-1)+d)}.$ We also transfer the obtained bounds to the total variation distance between the marginals of the estimated and the exactly guided samplers, and illustrate the performance of the suggested approach with numerical experiments.

Casting Bayesian inference as a neural network optimization problem targeting an amortized posterior is attractive, as it extends to otherwise intractable statistical models and offers near instantaneous inference for new datasets after prepaying the training cost. Although theory guarantees faithfulness under ideal convergence, practical amortized inference still requires iterating over architectures and optimization choices and ultimately “satisficing” under finite simulation, compute, and time budgets. Even the best-performing solution may thus retain avoidable representation gaps that typically require problem-specific fixes. Here, we propose a generic alternative which improves training dynamics with auxiliary guidance losses applied to internal representations. Specifically, we show how such guidance leads to faster convergence when training data is abundant and to better performance when it is scarce. We formalize representation gaps as getting stuck in a local optimum at the information bottleneck between the parts of the network tasked with feature learning and those tasked with conditional distribution learning, and offer a generic diagnostic to separate summary failures from inference failures. Finally, we demonstrate that auxiliary supervision improves convergence speed and accuracy on a range of challenging real-world inference problems.

We present B-GRASP (Bayesian GRAph inference with SPDE priors), a Bayesian framework for inferring uncertain edge weights in stochastic dynamical systems on graphs from noisy observations of nodal states. The unknown edge weights parameterize the graph differential operator and therefore directly govern the evolution of the graph process. Motivated by connections between differential operators in PDEs and SPDEs and their graph counterparts, we construct stochastic graph models incorporating diffusion, reaction dynamics, and stochastic forcing. The resulting hierarchical formulation jointly represents uncertainty in the graph structure and stochastic forcing. Latent graph variables determine positive edge weights and the corresponding graph Laplacian, while latent Brownian variables represent the stochastic forcing. Conditional on these variables, the graph dynamics define a deterministic forward map from which the likelihood and posterior distribution are constructed. We characterize the posterior using maximum a posteriori estimation and the No-U-Turn Sampler, enabling both point estimation and uncertainty quantification. We demonstrate the framework on a one-dimensional inverse heat-conduction problem, stationary and nonstationary graph reaction-diffusion systems with nonlinear dynamics, and state-level COVID-19 data in the United States. The numerical results show that posterior uncertainty provides information not captured by point estimates, particularly for weakly identifiable or highly conductive edges, and enables uncertainty in both graph connectivity and stochastic forcing to be quantified within a Bayesian framework.

Randomized smoothing certifies the probability of a fixed output event as the center of Gaussian noise moves. Feasibility or confidence filtering reports label probabilities only among retained proposals, producing a ratio. Its numerator is a fixed Gaussian event mass, while its denominator is the probability of retention and can change with the center. Substituting this ratio into the ordinary smoothing formula can therefore certify a ball that contains a decision boundary. We separate the problem into a geometric question and a certification question. Geometry determines when conditioning preserves Gaussian comparisons. Convex retained sets preserve the full comparison, while general sets require geometric control of the retained law as the center moves. Without such control, conditional probabilities imply no positive universal radius. Joint retention-and-label probabilities always yield a valid certificate for the same filtered predictor. A uniform covariance bound transfers divergence certificates to the retained law and can yield larger radii even when the Gaussian event comparison fails. Both methods admit finite-sample bounds. For a learned image classifier with a training-selected nonconvex filter, conditional Rényi bounds certify more images than joint-mass bounds without additional model evaluations. A released confidence filter exhibits verified label changes inside radii obtained by conditional substitution. An application of adaptive Gaussian composition covers causal finite-horizon executions with history-dependent center shifts under a pathwise energy bound.

Algorithms are increasingly being used to help automate and improve data-driven decisions, but care must be taken to prevent such algorithms from learning discriminatory patterns from historical data and perpetuating their biases. Statistical notions of fairness aim to mitigate either a model's disparate impact on disadvantaged groups (thus ensuring group fairness) or the resulting disparate treatment of individuals with similar features (thus ensuring individual fairness). Simultaneously mitigating disparate impact and disparate treatment is generally impossible for non-trivial models, necessitating a compromise. In this paper, we introduce the Fair Lasso and Fair Posterior as methods for selecting fair covariates in generalised linear models. By targeting variables that are simultaneously strong predictors of the response and weakly dependent on the sensitive group memberships, we aim to achieve favourable trade-offs between disparate treatment and disparate impact. Additionally, our selected set of fair features can be used as the conditioning set of legitimate features in the paradigm of Conditional Demographic parity (CDP) when no prescriptive legal framework exists.

Longitudinal clinical cohorts are multimodal, irregularly sampled and pervasively incomplete: in ADNI, positron emission tomography and cerebrospinal fluid assays are absent from roughly half of all visits. Linear state-space models handle irregular sampling gracefully but treat a missing modality by masking the input, leaving the transition operator untouched. We prove that this is a representational limitation: the latent state of any linear state-space layer whose transition operator does not depend on the availability pattern is an additive function of the availability indicators, so no such layer can represent an interaction between two modalities being jointly present or jointly absent. We propose CAMOS, which gives each modality a bank of second-order oscillators coupled through a matrix that sits inside the differential equation and is gated by availability, so the transition operator itself becomes a function of which measurements were taken. Coupling invalidates the analysis of uncoupled oscillatory models, and we restore it: a per-channel Gershgorin budget makes the effective stiffness positive definite uniformly over all $2^M$ availability patterns and all gaps, an energy argument charges amplification to availability transitions rather than sequence length, and a channel factorization preserves exact associative parallel scans. On ADNI, CAMOS outperforms uncoupled oscillatory state-space models and clinical fusion models on same-visit staging, landmark prediction and longitudinal forecasting, and under zero-shot transfer to OASIS-3 it is the only model that avoids collapse to the majority class.

Distributionally robust optimization (DRO) studies parameter estimation under uncertainty in the underlying probability distribution and has emerged as a principled framework for analyzing robustness and generalization. In particular, Wasserstein DRO, with distributional uncertainty induced by the Wasserstein distance, generalizes several popular regularizers. This paper studies Wasserstein DRO linear regression, unifying square-root Lasso and adversarial linear regression as important special cases. We prove that many properties of these two special cases carry over to this general method. In particular, we show (i) deterministic and non-asymptotic in-sample error bounds $O(n^{-1/2})$ in general and $O(n^{-1})$ under design matrix and sparsity conditions; (ii) insensitivity to the noise level, also known as the pivotal property; and (iii) solution equivalences for small and large ambiguity sets. The key proof step is to recast the method into a quadratic form, mimicking adversarial linear regression. We also show that the method can be solved efficiently, and we validate our findings through numerical simulations.

Time series foundation models (TSFMs) commonly adapt to new data by attaching a single trainable head to a frozen backbone, a one-size-fits-all setup that underfits heterogeneous regimes. Replacing the head with a mixture of experts is the standard upgrade, but on instance-normalized backbones (the dominant TSFM design class) it fails: routing entropy collapses to zero and one expert absorbs every input, a failure we call normalization-induced routing collapse. Standard MoE rescue mechanisms do not repair it, because the cause is in the router's input, not its optimization. Pre-encoder normalization strips the statistics a router would need to tell regimes apart. A mutual-information decomposition makes this precise and yields a signal-ratio that, computed before training, predicts dataset vulnerability (Spearman $\rho = -0.88$). Eight causal controls, including a vision-modality replication, isolate instance normalization as the cause. The prescription is a minimal causal intervention: Raw-Routed Mixture of Adapters (RR-MoA), which routes on the raw, pre-normalization input. Under a strictly frozen backbone, RR-MoA wins 54/54 comparisons against the strongest fixed adapter and significantly outperforms LoRA, TRACE, AdaMix, and full fine-tuning. The effect generalizes across six backbones and an imputation task. Frozen RR-MoA also beats full fine-tuning by 12-79% (the Frozen Paradox); two architecturally distinct variants confirm the principle generalizes beyond this specific router.

We compare the instance-wise, finite-sample risks of monotone spectral filters for linear regression, a broad class of estimators including principal component regression (PCR), gradient descent (GD), and ridge regression. We show that PCR dominates all monotone spectral filters: compared to any such filter, the risk of optimally tuned PCR is no bigger by a constant factor for all problems. Furthermore, the dominance is strong if the filter is separated from step functions (e.g., GD and ridge): there exist problem instances for which the risk of PCR is smaller by a polynomial factor in sample size dependence. Our comparison results show that PCR is optimal and thus admissible among monotone filters, significantly extending Wu et al. (2026)'s result that GD strongly dominates ridge. From a technical perspective, we establish new upper and lower bounds for general spectral filters, which are instance-wise sharp when specialized to ridge or GD, recovering or improving the best-known bounds.

Population loss can remain nearly constant while a neural network learns a substantially more predictive representation. We establish this separation for two-layer ReLU and leaky-ReLU networks trained on Gaussian inputs by simultaneous fixed-step population gradient descent on all parameters. For structured additive teachers whose links are positive mixtures of Gaussian-damped cubics in $H^1(\gamma)$, we give explicit conditions under which small IID Gaussian initialization yields a high-probability guarantee: at a checkpoint during a high-loss plateau, minimum alignment between the rank-$r$ teacher subspace and the leading $r$-dimensional eigenspace of the predictor's average gradient outer product (AGOP) increases by at least $1/2$, and the minimum refit MSE under unchanged coefficient budgets decreases by more than $0.399$, both relative to initialization. The same trajectory subsequently attains a trained loss below every value in the plateau window. A complementary result treats unequal-weight cubic teachers and small additive Sobolev perturbations using projected-feature refits. For SwiGLU networks with an exactly fitted intercept, we prove leading-AGOP alignment during a loss plateau at fixed width and dimension as Gaussian initialization vanishes, for square-integrable teachers with nonzero Hermite content of degree one, two, or three. A rank-one cubic specialization also gives simultaneous unrestricted-refit gains at a prescribed width. An approximation lower bound further shows that certain interaction targets retain nonzero error when ridge neurons are restricted to shared orthogonal axes within the teacher subspace. Population-moment experiments with ReLU students across 21 teachers and 50 initializations per teacher complement the analysis.

We study data-driven multi-period lost-sales inventory control under censored demand, where a stockout reveals only that demand exceeded the stocking level. We develop a unified, model-based framework for policy learning from censored data, built on a new cost decomposition for base-stock policies and a biased sample-average approximation (SAA) approach. The cost decomposition allows us to propose a new coverage condition under which censored observations are informative enough for sample-efficient policy learning. Guided by this coverage condition, we design two biased SAA algorithms: an upper-biased one that achieves near-optimal sample complexity under the offline coverage condition, and a lower-biased one that actively generates the required coverage and achieves near-optimal regret online. More broadly, this biased SAA approach provides a general principle for implementing pessimism and optimism under censored feedback, which may be of independent interest.

Despite the widespread use of Low-Rank Adaptation (LoRA), little is known about its dynamics in continual learning and the mechanisms by which low-rank updates affect catastrophic forgetting. We provide an asymptotically exact dynamical characterization of LoRA in a solvable two-task teacher-student model. In the high-dimensional online-learning limit, we derive a closed system of ordinary differential equations for a finite set of macroscopic order parameters, yielding exact expressions for the generalization errors throughout both the initial Task 1 learning phase and the subsequent LoRA fine-tuning on Task 2. The theory quantitatively matches finite-dimensional simulations and exposes two characteristic effects of LoRA: low-rank adaptation reduces interference with features learned on the first task, but its initialization slows adaptation to the second task. Building on this mechanistic picture, we analyze a state-dependent masking strategy that freezes hidden units carrying the strongest first-task representations and restricts adaptation to the complementary subspace. This structural partitioning markedly reduces forgetting, while preserving plasticity on the new task. Our framework further clarifies the role of adapter rank: transfer improves only up to the intrinsic dimensionality of the target task and saturates beyond it, while forgetting continues to grow with rank. These results provide a dynamical and geometric account of how low-rank adaptation organizes information across sequential tasks and are qualitatively reproduced on a sequential MNIST benchmark.

We develop a general second-order asymptotic framework for estimators constructed from invariant probability measures of data-dependent Markov processes. The framework yields a universal bias correction for smooth functionals, computed from the same invariant measure used to construct the estimator, without the resampling and repeated estimation involved in bias-correction methods such as the bootstrap and jackknife. The statistical input is represented by a random one-form, allowing differentials of log-likelihoods and general random criteria, as well as estimating functions, to be treated in a common framework. Under local regularity and localization conditions, we derive second-order stochastic expansions for averages with respect to these invariant measures and for decision rules defined through general loss functions. The expansions separate the contributions of the statistical input, temperature, drift, and loss, using precontrast geometry and an O-derivative induced by the loss. The framework includes maximum likelihood and posterior-based estimators and recovers existing bias-reduction methods based on the choice of prior or adjustments to estimating equations. In the likelihood setting, for the Jeffreys posterior $\mu_n$ and its Fisher–Rao Fréchet mean $\widehat z_n$, the corrected estimator $2\gamma(\widehat z_n)-\mu_n(\gamma)$ has frequentist bias $o(a_n^2)$ for every fixed smooth functional $\gamma$, where $a_n$ denotes the estimation rate. For coordinate squared loss, the theory yields a diffusion determined only by the likelihood and Fisher information. The mean of its invariant probability measure estimates the parameter with bias $o(a_n^2)$. This construction remains available even when no bias-reducing prior exists. Numerical results in the gamma shape–scale model illustrate this bias reduction.

Matching problems are ubiquitous in data science as they enable the alignment of structured objects and distributions. While existing solvers are often tailored to specific matching formulations, we unify a broad class of such problems within a common mathematical and optimization framework based on duality theory. Theoretically, we demonstrate that matching objectives decomposable as a difference of convex (DC) functions can be recast as implicit registration problems. This connection links matching to another well-studied class of objectives and yields a dual formulation amenable to natural optimization strategies. We then apply these findings to quadratic matching (QM) problems, which admit DC decompositions and for which we provide extensive convergence guarantees. Our framework applies to Gromov-Wasserstein (GW), as well as its unbalanced formulation and several variants, which are increasingly popular QM problems. Numerically, we implement our algorithms at scale for various data modalities such as graphs, point clouds, meshes, and word embeddings. Finally, our modular approach allows us to explore new formulations such as fracture matching, broadening the scope of problems that can be addressed within this framework..

Count data pose a challenge for score-matching-based causal discovery: derivatives are unavailable, and simply replacing them with finite differences does not generally suffice for causal discovery. We generalize SCORE's constant-curvature criterion (Rolland et al., 2022) by conditioning on the node's value, yielding the conditional curvature score (CCS) for ordering. We also extend curvature-based parent recovery through the off-diagonal curvature score (OCS), enabling directed acyclic graph (DAG) recovery with both scores constructed from score functions for continuous data and concrete scores for counts. In the bivariate setting, zero CCS exactly characterizes a semiparametric generalized linear model (GLM) conditional form in which the conditional family need not be specified in advance, unlike in classical GLMs. For bivariate semiparametric GLM DAGs under our regularity condition, canonical-parameter nonlinearity is necessary and sufficient for identifiability. In multivariate DAGs, this nonlinearity enables DAG recovery through CCS and OCS. Our framework identifies a new class of semiparametric GLM DAGs that strictly contains the nonlinear Gaussian ANM class identified by SCORE. We introduce DISCO (DIscrete SCOre), a count-DAG recovery algorithm that estimates CCS and OCS using discrete diffusion. Experiments demonstrate accurate DAG recovery across Poisson, negative binomial, binomial, and mixed-family settings, as well as scalability to 1,000-node DAGs on a single GPU.

Reinforcement learning and contextual bandit algorithms have become increasingly common in sequential decision-making applications. When these methods are deployed in high-stakes domains, there is growing interest not only in learning effective policies, but also in conducting statistical inference for quantities learned under adaptive data collection. However, classical procedures applied naively in these settings can fail: even when estimators are unbiased, their variance becomes path-dependent and as a result may not be asymptotically normal. A growing literature has emerged to ameliorate this problem, but solutions tend to be problem specific and often rely on correct specification of a working model. In this work, we develop a unified framework for constructing asymptotically valid confidence intervals to cover smooth functionals of nonparametric M-estimands under adaptive sampling. Under Neyman orthogonality, we provide two novel methods for performing inference: (1) a self-normalized statistic based on the realized quadratic variation of the influence function and (2) a statistic using a plug-in estimate of the conditional variance based on reweighted influence function increments. Our results allow for flexible nonparametric estimation of nuisance parameters and remain valid under model misspecification. Our theory is supported by a simulation study for a dynamic pricing application which demonstrates that this method can produce asymptotically valid confidence intervals where standard methods fail.

We study additive regression under an unknown and potentially non product design distribution, allowing the number of covariates to grow with the sample size. We consider a coupled class that separately controls the smoothness of the marginal densities and of each additive component multiplied by the corresponding marginal density. Under joint-density bounds that hold uniformly in the dimension and suitable dimension-growth conditions, we establish matching minimax bounds for prediction. With known marginal densities, the classical additive rate is attainable. When the marginals are unknown, this rate is preserved if the densities are at least as smooth as the weighted components. Otherwise, marginal-density smoothness determines the minimax rate over the coupled class. Finally, we recover all additive components with total squared error of the same order as the prediction error.

In order to estimate the longitudinal trend, we often use the generalized multivariate analysis of variance (GMANOVA) model (Potthoff & Roy, 1964), when the longitudinal data is balanced data. Usually, we use this GMANOVA model with some polynomial at the time of measurement. However, when the longitudinal trend has flexible curve, we cannot derive good fitting estimated curve when we use some polynomial curves. Then, Nagai (2011) proposed the nonparametric GMANOVA model which uses on several known basis functions instead of using the polynomial curves. If we use several basis functions, then overfitting problem is occurred. Nagai (2011) also proposed the estimation method for avoiding overfitting and unstable problems, and reducing computational iterative algorithm by extending the generalized ridge regression model (Yanagihara, Nagai & Satoh, 2009). In the present paper, we extend one of the optimization methods in Nagai, Yanagihara and Satoh (2012) into the estimation method in Nagai (2011). Through numerical studies, we show some properties of each optimization method.

In this paper, we consider asymptotic properties of the support vector machine (SVM) in high-dimension, low-sample-size (HDLSS) settings under a spiked model. The existing theory of the SVM in the HDLSS context relies on the geometric representation of HDLSS data, which requires that the eigenvalues of the covariance matrices are not dominant. We first show that the geometric representation does not hold under the spiked model. We show that the Gram matrix of HDLSS data converges in distribution to a random matrix, namely, the HDLSS data converge to a random configuration in a finite-dimensional space whose dimension is given by the number of the spikes. We show that the misclassification rates of the SVM do not tend to zero, that is, the SVM does not hold the consistency property. We also show that the bias-corrected SVM (BC-SVM) does not give preferable performance in this setting because the bias term itself should be modified. In order to overcome such difficulties, we propose a spike-corrected SVM (SC-SVM). We show that the SC-SVM holds the consistency property when the sample size goes to infinity, and that the growth of the sample size is essential in the sense that any projection-based procedure fails when the sample size is fixed. Finally, we check the performance of the classifiers by numerical simulations.

Inference-time alignment of flow and diffusion-based models is critical for achieving flexible generative modeling. Theoretically, Doob's $h$-transform provides an elegant solution to this problem, and most existing methods are based on this principle. However, in practice, estimating the optimal guidance derived from Doob's $h$-transform at inference time is challenging. To deal with this issue, we regard inference-time alignment as a sequential optimization problem in the space of probability measures and propose a novel framework called *Steepest Guidance*, based on the principle of maximizing local improvement in the objective. We provide a theoretical analysis of the proposed method and demonstrate its effectiveness through extensive experiments.

Regression analysis for stochastic processes has been an important topic over the last decades. Typically, regressing a dependent process $Y$ on an explanatory $Z$ results in an integral relationship of the form $dY_t = \beta_t dZ_t + dX_t$ with some residual process $X$, and often with constant or piecewise constant $\beta$. In the case of jump processes such a relationship essentially boils down to $\Delta Y_t = \beta \Delta Z_t + \Delta X_t$ on the level of the jumps. Here the residual process $X$ is such that it never jumps together with $Z$, and so any jump in $Y$ is either exactly proportional to $Z$ or purely idiosyncratic. In this paper we discuss a related setup which appears more realistic and is closer to classical regression, namely $\Delta Y_t = (\beta + \eta_t) \Delta Z_t$ for i.i.d. $(\eta_t)_{t \ge 0}$. We propose an estimator for $\beta$ based on high-frequency observations of $(Y,Z)$ and discuss consistency and associated central limit theorems in two different asymptotic regimes.

Fairness assessment in algorithmic decisions that affect individuals, such as credit scoring, often relies on parity measures calculated at the aggregate group level. Such measures may not reveal which individuals experience unfairness or which explanatory factors contribute to it. In this paper, we propose a rank-based framework that evaluates fairness through the distribution of model prediction errors, thereby linking fairness assessment with predictive accuracy and explainability. The framework combines Rank Graduation Fairness (RGF), its integrated measure AURGF, a centered Cramer–von Mises permutation test, and a feature removal procedure for fairness explainability. We evaluate the methodology using logistic regression, random forest, gradient boosting, and a multilayer perceptron. The simulation study shows that protected-group imbalance can reverse descriptive fairness comparisons, whereas the proposed inferential procedure correctly distinguishes fair from unfair mechanisms. Its application to HMDA mortgage data produces model rankings that differ from those obtained with classical fairness criteria. Tree-based models, rather than logistic regression, provide the strongest combination of predictive accuracy and rank-based fairness, while the fairness null hypothesis is rejected for all four models. The persistence of disparity across statistical, bagging, boosting, and neural network specifications, together with the feature removal results, indicates that the observed unfairness is not specific to a single algorithm or predictor, but is associated with group differences embedded in the characteristics of the lending data. These findings support a broader approach to trustworthy artificial intelligence that combines predictive accuracy, fairness measurement, statistical inference, and explainability.

We study how well deep neural networks approximate and learn spectral Barron functions. Recent studies have shown that these function classes can be efficiently approximated by shallow neural networks without suffering from the curse of dimensionality. We complement these results by providing new approximation bounds for deep networks with ReLU activation and establishing the minimax rates for learning these function classes. Specifically, we show that $d$-dimensional spectral Barron functions with smoothness index $s>0$ can be approximated by deep ReLU neural networks with approximation rate $\widetilde{\mathcal{O}} (S^{-\frac{1}{2}-\frac{s}{d}})$, where $S$ denotes the number of nonzero parameters in the network. Using this approximation result, we further show that deep ReLU neural networks can learn spectral Barron functions in a fast rate $n^{-\frac{d+2s}{2d+2s}}$ with $n$ training samples. Finally, we prove that this convergence rate is minimax optimal up to logarithmic factors.