2026 年的期刊文章 · 第 44 页

共 1756 篇,按发表时间排,新的在前,每页 40 篇。这一页是 2026 到 2026。

按期刊、按主题看 →

第 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 页

JMLR Optimization & Learning Theory · 发表于 2026

Approximations and Learning for Continuous State and Action MDPs under Average Cost Criteria

Ali D. Kara, Serdar Yüksel

摘要

In this paper, for Markov Decision Processes (MDPs) with standard Borel spaces, (i) we first provide a discretization based approximation method for MDPs with continuous spaces under average cost criteria, and provide error bounds for approximations when the dynamics are only weakly continuous (for asymptotic convergence of errors as the grid sizes vanish) or Wasserstein continuous (with a rate in approximation as the grid sizes vanish) under certain ergodicity assumptions. In particular, we relax the total variation condition given in prior work to weak continuity or Wasserstein continuity. (ii) We provide synchronous and asynchronous (quantized) Q-learning algorithms for continuous spaces via quantization (where the quantized state is taken to be the actual state in corresponding Q-learning algorithms presented in the paper), and establish their convergence. (iii) We finally show that the convergence is to the optimal Q values of a finite approximate model constructed via quantization, which implies near optimality of the arrived solution.

JMLR Optimization & Learning Theory · 发表于 2026

Approximation-Free Differentiable Oblique Decision Trees

Subrat Prasad Panda, Blaise Genest, Arvind Easwaran

摘要

Decision Trees (DTs) are widely used in safety-critical domains such as medical diagnosis, valued for their interpretability and effectiveness on tabular data. However, training accurate oblique DTs is challenging due to complex optimization landscapes and overfitting risks, particularly in regression. Recent advances have introduced differentiable formulations that enable gradient-based training and joint optimization of decision boundaries and leaf regressors. Yet, existing approaches typically rely on approximations, either through probabilistic softening of boundaries (soft DTs) or quantized gradients such as the Straight-Through Estimator (STE). To overcome these limitations, we propose DTSemNet, a novel, semantically equivalent, and invertible representation of hard oblique DTs as neural networks. DTSemNet enables end-to-end training with standard gradient descent, eliminating the need for approximations in both classification and regression. While classification aligns naturally with this formulation, regression remains challenging due to the joint optimization of internal nodes and leaf regressors. To address this, we analyze the limitations of STE and introduce an annealed Top-$k$ method that provides accurate gradient signals without approximation. Extensive experiments on classification and regression benchmarks show that DTSemNet-trained oblique DTs outperform state-of-the-art differentiable DTs. Furthermore, we demonstrate that DTSemNet can serve as programmatic DT policies in reinforcement learning environments, thereby broadening their applicability.

JMLR Optimization & Learning Theory · 发表于 2026,收录于 2026-11-06

Anderson Mixing in Bures–Wasserstein Space of Gaussian Measures

Vitalii Aksenov, Martin Eigel, Mathias Oster

摘要

Various statistical tasks, including sampling or computing Wasserstein barycenters, can be reformulated as fixed-point problems for operators on probability distributions. Accelerating standard fixed-point iteration schemes provides a promising novel approach to the design of efficient numerical methods for these problems. The Wasserstein geometry on the space of probability measures, although not precisely Riemannian, allows us to define various useful Riemannian notions, such as tangent spaces, exponential maps and parallel transport, motivating the adaptation of Riemannian numerical methods. We demonstrate this by developing and implementing the Riemannian Anderson mixing (RAM) method for Gaussian distributions. The method reuses the history of the residuals and improves the iteration complexity, and we argue that the additional costs, compared to the Picard method, are negligible. We show that certain balls in the Bures--Wasserstein manifold satisfy the requirements for convergence of RAM. The numerical experiments show a significant acceleration compared to the Picard iteration, and performance on par with Riemannian gradient descent and conjugate gradient methods.

Econometrica Econometrics & Economics · 发表于 2026

An Experimental Evaluation of Deferred Acceptance: Evidence From Over 100 Army Officer Labor Markets

Jonathan M.V. Davis, Kyle Greenberg, Damon Jones

摘要

Internal labor markets are increasingly important for matching workers to jobs within organizations. We present evidence from a randomized trial that compares matching workers to jobs using the deferred acceptance (DA) algorithm to the traditional manager‐directed matching process. Our setting is the U.S. Army's internal labor market, which matches over 14,000 officers to units annually. We find that DA reduces administrative burden and increases match quality as measured by reduced justified envy, increased truthful preference reporting, and officers' and units' preferences over their matches. The overall impact of DA on officer retention and performance in the two years after officers started their new jobs is limited by strategic preference coordination between officers and units. However, DA leads to significant improvements in officer retention and promotions in markets with inexperienced managers. Our findings suggest that cross‐market communication between agents in internal labor markets can attenuate the benefits of strategyproof matching algorithms.

JMLR Optimization & Learning Theory · 发表于 2026,收录于 2026-11-06

An Efficient HPR Algorithm for the Wasserstein Barycenter Problem with O(Dim(P)/ε) Computational Complexity

Guojun Zhang, Yancheng Yuan, Defeng Sun

摘要

In this paper, we study and implement an efficient Halpern--Peaceman--Rachford (HPR) algorithm for solving the Wasserstein barycenter problem (WBP) with fixed supports. While the Peaceman--Rachford (PR) splitting method itself may not converge for the WBP, applying the Halpern iteration to the PR operator yields an $O(1/\varepsilon)$ non-ergodic iteration complexity with respect to the Karush--Kuhn--Tucker (KKT) residual. The structured linear solve in the HPR subproblem corresponds to an affine projection that can be viewed as a coupled extension of Romero's classical OT projection formula. This observation yields a closed-form solution procedure with cost linear in the size of the linear system. Consequently, HPR achieves an $O({\rm Dim(P)}/\varepsilon)$ computational complexity in terms of flops for obtaining an $\varepsilon$-optimal solution measured by the KKT residual of the unregularized WBP, where ${\rm Dim(P)}$ denotes the number of variables in the corresponding LP formulation. This bound matches the order of the best-known bound for the WBP, while requiring no entropic regularization and measuring accuracy by the KKT residual. Numerical experiments on synthetic and real data sets demonstrate the practical efficiency of the restarted HPR implementation for large-scale WBPs.

JMLR Other · 发表于 2026

An Anytime Algorithm for Good Arm Identification

Marc Jourdan, Andrée Delahaye-Duriez, Clémence Réda

摘要

In good arm identification (GAI), the goal is to identify one arm whose average performance exceeds a given threshold, referred to as a good arm, if it exists. Few works have studied GAI in the fixed-budget setting when the sampling budget is fixed beforehand, or in the anytime setting, when a recommendation can be asked at any time. We propose APGAI, an anytime and parameter-free sampling rule for GAI in stochastic bandits. APGAI can be straightforwardly used in fixed-confidence and fixed-budget settings. First, we derive upper bounds on its probability of error at any time. They show that adaptive strategies can be more efficient in detecting the absence of good arms than uniform sampling in several diverse instances. Second, when APGAI is combined with a stopping rule, we prove upper bounds on the expected sampling complexity, holding at any confidence level. Finally, we show the good empirical performance of APGAI on synthetic and real-world data. Our work offers an extensive overview of the GAI problem in all settings.

JMLR Optimization & Learning Theory · 发表于 2026,收录于 2026-11-06

Alternating Regret for Online Convex Optimization

Soumita Hait, Ping Li, Haipeng Luo, Mengxiao Zhang

摘要

Motivated by alternating learning dynamics in two-player games, a recent work by Cevher et al. (2023) shows that $o(\sqrt{T})$ alternating regret is possible for any $T$-round adversarial Online Linear Optimization (OLO) problem, and left as an open question whether the same is true for general Online Convex Optimization (OCO). We answer this question in the affirmative by showing that the continuous Hedge algorithm achieves $\tilde{\mathcal{O}}(d^{\frac{2}{3}}T^{\frac{1}{3}})$ alternating regret for any adversarial $d$-dimensional OCO problems. We show that this implies an alternating learning dynamic that finds a Nash equilibrium for any convex-concave zero-sum games or a coarse correlated equilibrium for any convex two-player general-sum games at a rate of $\tilde{\mathcal{O}}(d^{\frac{2}{3}}/T^{\frac{2}{3}})$. To further improve the time complexity and/or the dimension dependence, we propose another simple algorithm, Follow-the-Regularized-Leader with a regularizer whose convex conjugate is 3rd-order smooth, for OCO with smooth and self-concordant loss functions (such as linear or quadratic losses). We instantiate our algorithm with different regularizers and show that, for example, when the decision set is the $\ell_2$ ball, our algorithm achieves $\tilde{\mathcal{O}}(T^{\frac{2}{5}})$ alternating regret with no dimension dependence (and a better $\tilde{\mathcal{O}}(T^{\frac{1}{3}})$ bound for quadratic losses). We complement our results by showing some algorithm-specific alternating regret lower bounds, including a somewhat surprising $\Omega(\sqrt{T})$ lower bound for a Regret Matching variant that is widely used in alternating learning dynamics.

JMLR Other · 发表于 2026

Almost Sure Convergence of Linear Temporal Difference Learning with Arbitrary Features

Jiuqi Wang, Shangtong Zhang

摘要

Temporal difference (TD) learning with linear function approximation (linear TD) is a classic and powerful prediction algorithm in reinforcement learning. While it is well-understood that linear TD converges almost surely to a unique point, this convergence traditionally requires the assumption that the features used by the approximator are linearly independent. However, this linear independence assumption does not hold in many practical scenarios. This work is the first to establish the almost sure convergence of linear TD without requiring linearly independent features. We prove that the weight iterates of linear TD converge to a bounded set, and that the value estimates derived from the weights in that set are the same almost everywhere. We also establish a notion of local stability of the weight iterates. Importantly, we do not impose assumptions tailored to feature dependence and do not modify the linear TD algorithm. Key to our analysis is a novel characterization of bounded invariant sets of the mean ODE of linear TD.

JMLR Machine Learning Networks & Graphs · 发表于 2026

AgentPEN: A Prediction-Explanation Network for Sequential Stock Movement via LLMs and Recurrent Generation

Shuqi Li, Mengyao Guo, Yunzhong Zheng, Siqi Li, Xin Gao, Rui Yan

摘要

The importance of explainability in stock prediction is increasingly recognized, especially for audit and regulatory purposes. Meanwhile, financial news corpora are often key drivers behind stock price fluctuations. However, the raw news data obtained is usually highly noisy, has a highly variable scope of influence in time and space, and is not precisely synchronized with stock price data. In this paper, we propose a prediction-explanation network called AgentPEN, which can provide clear explanations for complex temporal price patterns. Specifically, AgentPEN jointly aligns text and price streams by an LLM-based Representation Fusion Agent and then adopts a Deep Recurrent Generation module to explore the distribution of stock movements. The LLM-based Representation Fusion Agent is designed in a Selection-Memory-Fusion manner: the Text Selection Module picks up useful information from massive text data; the Text Memory Module evaluates and writes the text memory from a two-view perspective, including Temporal Memory and Spatial Memory; the Information Fusion Module models the interaction between text and price data. Next, the fused representation is sent to the Deep Recurrent Generation module to convert insights into stock movement predictions. Experiments on multiple real-world datasets have shown that AgentPEN surpasses the state-of-the-art baselines both in prediction accuracy and explainability.

JMLR Machine Learning Networks & Graphs · 发表于 2026

Adversarial Rademacher Complexity of Deep Neural Networks

Jiancong Xiao, Yanbo Fan, Ruoyu Sun, Zhi-Quan Luo

摘要

Deep neural networks (DNNs) are highly vulnerable to adversarial attacks. Ideally, a robust model should perform well on both perturbed training data and unseen perturbed test data. While DNNs can fit perturbed training data, generalizing to perturbed test data remains a significant challenge. This motivates the study of generalization guarantees from a learning theory perspective. This paper focuses on adversarial Rademacher complexity (ARC), first introduced by Khim and Loh (2018) and Yin et al. (2019). Their work primarily addressed linear functions and highlighted the open question of how to bound ARC for neural networks. Since then, several attempts have been made, with the latest results applying ARC only to two-layer neural networks. The main challenge arises from the dynamic nature and unknown closed-form solution of adversarial examples. In this paper, we resolve this issue and provide the first bound on ARC for deep neural networks. Our bound is qualitatively comparable to Rademacher complexity bounds in similar settings. The key ingredient is a new concept we introduce, termed intermediate adversarial examples, along with a framework for calculating the covering number that is compatible with them. Finally, we present experiments to analyze poor robust generalization, demonstrating that the weight norm is a crucial factor influencing the robust generalization gap.

JMLR Bayesian Methods Nonparametric & Semiparametric · 发表于 2026

Adaptive Nonparametric Perturbations of Parametric Models with Generalized Bayes

Bohan Wu, Eli N. Weinstein, Sohrab Salehi, Yixin Wang, David M. Blei

摘要

Parametric Bayesian modeling offers a powerful and flexible toolbox for machine learning. Yet the model, however detailed, may still be wrong, and this can make inferences untrustworthy. In this paper we introduce a new class of semiparametric corrections for parametric Bayesian models, when the target of inference is a functional of the true data distribution. Our starting point is a fully Bayesian modeling approach, which explicitly accounts for the possibility that the parametric model is wrong. Asymptotic analysis shows that this approach is both robust to model misspecification and data efficient, achieving fast convergence when the parametric model is close to true. However, the fully Bayesian approach is limited in its practical usefulness by the challenges of conducting inference and computing a Bayes factor for a nonparametric model. We therefore propose a novel model correction based on generalized Bayes, which entirely avoids the need to compute a nonparametric Bayes factor, but preserves the robustness and efficiency of the fully Bayesian approach. We demonstrate our method by estimating causal effects of gene expression from single cell RNA sequencing data. Overall, we offer a new efficient approach to robust Bayesian inference with parametric models.

JMLR High-Dimensional · 发表于 2026

Adaptive Forward Stepwise: A Method for High Sparsity Regression

Ivy Zhang, Robert Tibshirani

摘要

This paper proposes a sparse regression method that continuously interpolates between Forward Stepwise selection (FS) and the LASSO. When tuned appropriately, our solutions are much sparser than typical LASSO fits but, unlike FS fits, benefit from the stabilizing effect of shrinkage. Our method, Adaptive Forward Stepwise Regression (AFS) addresses the need for sparser models with shrinkage. We show its connection with boosting via a soft-thresholding viewpoint and demonstrate the ease of adapting the method to classification tasks. In both simulations and real data, our method has lower mean squared error and fewer selected features across multiple settings compared to popular sparse modeling procedures.

JMLR Optimization & Learning Theory · 发表于 2026,收录于 2026-09-10

Adaptive Algorithms for Infinitely Many-Armed Bandits: A Unified Framework

Emmanuel Pilliat

摘要

We consider a bandit problem where the budget is smaller than the number of arms, which may be infinite. In this regime, the usual objective in the literature is to minimize simple regret. To analyze broad classes of distributions with potentially unbounded support, where simple regret may not be well-defined, we take a slightly different approach and seek to maximize the expected simple reward of the recommended arm, providing anytime guarantees. To that end, we introduce a distribution-free algorithm, OSE, that adapts to the distribution of arm means and achieves near-optimal rates for several distribution classes. We characterize the sample complexity through the rank-corrected inverse squared gap function. In particular, we recover known upper bounds and transition regimes for $\alpha$ less or greater than $1/2$ when the quantile function is $\lambda_\eta = 1-\eta^{\alpha}$. We additionally identify new transition regimes depending on the noise level relative to $\alpha$, which we conjecture to be nearly optimal. Additionally, we introduce an enhanced practical version, PROSE, that achieves state-of-the-art empirical performance for the main distribution classes considered in the literature.

JMLR Other · 发表于 2026

Accelerating Constrained Sampling: A Large Deviations Approach

Yingli Wang, Changwei Tu, Xiaoyu Wang, Lingjiong Zhu

摘要

The problem of sampling a target probability distribution on a constrained domain arises in many applications including machine learning. For constrained sampling, various Langevin algorithms such as projected Langevin Monte Carlo (PLMC), based on the discretization of reflected Langevin dynamics (RLD) and more generally skew-reflected non-reversible Langevin Monte Carlo (SRNLMC), based on the discretization of skew-reflected non-reversible Langevin dynamics (SRNLD), have been proposed and studied in the literature. This work focuses on the long-time behavior of SRNLD, where a skew-symmetric matrix is added to RLD. Although acceleration for SRNLD has been studied, it is not clear how one should design the skew-symmetric matrix in the dynamics to achieve good performance in practice. We establish a large deviation principle (LDP) for the empirical measure of SRNLD when the skew-symmetric matrix is chosen such that its product with the outward unit normal vector field on the boundary is zero. By explicitly characterizing the rate functions, we show that this choice of the skew-symmetric matrix accelerates the convergence to the target distribution compared to RLD and reduces the asymptotic variance. Numerical experiments for SRNLMC based on the proposed skew-symmetric matrix show superior performance, which validate the theoretical findings from the large deviations theory.

JMLR Other · 发表于 2026

Abstract Gradient Training: A Unified Certification Framework for Data Poisoning, Unlearning, and Differential Privacy

Philip Sosnin, Matthew Wicker, Josh Collyer, Calvin Tsay

摘要

The impact of inference-time data perturbation (e.g., adversarial attacks) has been extensively studied in machine learning, leading to well-established certification techniques for adversarial robustness. In contrast, certifying models against training data perturbations remains a relatively under-explored area. These perturbations can arise in three critical contexts: adversarial data poisoning, where an adversary manipulates training samples to corrupt model performance; machine unlearning, which requires certifying model behavior under the removal of specific training data; and differential privacy, where guarantees must be given with respect to substituting individual data points. This work introduces Abstract Gradient Training (AGT), a unified framework for certifying robustness of a given model and training procedure to training data perturbations, including bounded perturbations, the removal of data points, and the addition of new samples. By bounding the reachable set of parameters, i.e., establishing provable parameter-space bounds, AGT provides a formal approach to analyzing the behavior of models trained via first-order optimization methods.

JMLR Causal Inference High-Dimensional · 发表于 2026

A causal fused lasso for interpretable heterogeneous treatment effects estimation

Oscar Hernan Madrid Padilla, Yanzhen Chen, Carlos Misael Madrid Padilla, Gabriel Ruiz

摘要

We propose a novel method for estimating heterogeneous treatment effects based on the fused lasso. By first ordering samples based on the propensity or prognostic score, we match units from the treatment and control groups. We then run the fused lasso to obtain piecewise constant treatment effects with respect to the ordering defined by the score. Similar to the existing methods based on discretizing the score, our methods yield interpretable subgroup effects. However, existing methods fixed the subgroup a priori, but our causal fused lasso forms data-adaptive subgroups. We show that the estimator consistently estimates the treatment effects conditional on the score under very general conditions on the covariates and treatment. We demonstrate the performance of our procedure using extensive experiments that show that it can be interpretable and competitive with state-of-the-art methods.

JMLR Machine Learning Econometrics & Economics · 发表于 2026

A Unified Approach to Analysis and Design of Denoising Markov Models

Yinuo Ren, Grant M. Rotskoff, Lexing Ying

摘要

Probabilistic generative models based on measure transport, such as diffusion and flow-based models, are often formulated in the language of Markovian stochastic dynamics, where the choice of the underlying process impacts both algorithmic design choices and theoretical analysis. In this paper, we aim to establish a rigorous mathematical foundation for denoising Markov models, a broad class of generative models that postulate a forward process transitioning from the target distribution to a simple, easy-to-sample distribution, alongside a backward process particularly constructed to enable efficient sampling in the reverse direction. Leveraging deep connections with nonequilibrium statistical mechanics and generalized Doob's $h$-transform, we propose a minimal set of assumptions that ensure: (1) explicit construction of the backward generator, (2) a unified variational objective directly minimizing the measure transport discrepancy, and (3) adaptations of the classical score-matching approach across diverse dynamics. Our framework unifies existing formulations of continuous and discrete diffusion models, identifies the most general form of denoising Markov models under certain regularity assumptions on forward generators, and provides a systematic recipe for designing denoising Markov models driven by arbitrary Lévy-type processes. We illustrate the versatility and practical effectiveness of our approach through novel denoising Markov models employing geometric Brownian motion and jump processes as forward dynamics, highlighting the framework's potential flexibility and capability in modeling complex distributions.

JMLR Machine Learning Optimization & Learning Theory · 发表于 2026

A Two-Timescale Primal-Dual Framework for Reinforcement Learning via Online Dual Variable Guidance

Axel F. Wolter, Tobias Sutter

摘要

We study reinforcement learning by combining recent advances in regularized linear programming formulations with the classical theory of stochastic approximation. Motivated by the challenge of designing algorithms that leverage off-policy data while maintaining on-policy exploration, we propose PGDA-RL, a novel primal-dual projected gradient descent-ascent algorithm for solving regularized Markov decision processes (MDPs). PGDA-RL integrates experience replay-based gradient estimation with a two-timescale decomposition of the underlying nested optimization problem. The algorithm operates asynchronously, interacts with the environment through a single trajectory of correlated data, and updates its policy online in response to the dual variable associated with the occupancy measure of the underlying MDP. We prove that PGDA-RL converges almost surely to the optimal value function and policy of the regularized MDP. Our convergence analysis relies on tools from stochastic approximation theory and holds under weaker assumptions than those required by existing primal-dual RL approaches, notably removing the need for a simulator or a fixed behavioral policy. Under a strengthened ergodicity assumption on the underlying Markov chain, we establish a last-iterate finite-time guarantee with $\widetilde{\mathcal O}(k^{-2/3})$ mean-square convergence, aligning with the best-known rates for two-timescale stochastic approximation methods under Markovian sampling and biased gradient estimates.

JMLR Other · 发表于 2026,收录于 2026-09-17

A Theoretical Framework for Masked Pretraining (MPT)

Qi Zhang, Runyu Zhou, Yifei Wang, Yisen Wang

摘要

Recently, Masked Pretraining (MPT) based on reconstruction pretraining tasks has risen to a promising self-supervised learning paradigm across various domains and achieves remarkable performance in multiple downstream tasks. However, the theoretical understanding of the working mechanism behind MPT is still limited. In this paper, we introduce a new theoretical framework to analyze MPT and understand the crucial role of masking in extracting meaningful representations. We establish theoretical connections between MPT and another popular self-supervised paradigm: contrastive learning. We prove that the masking technique implicitly creates positive pairs that are semantically similar and the reconstruction loss pulls them together in the feature space. Besides, as a result of the implicit alignment, we point out the dimensional collapse issue of MPT and propose a Uniformity-enhanced MPT (U-MPT) loss that can effectively address this issue and bring significant improvements in downstream tasks including linear evaluation, cross-dataset fine-tuning and out-of-distribution generalization on real-world data sets. Furthermore, we establish downstream guarantees of U-MPT and theoretically analyze the influence of masking strategies. Based on the theoretical analysis, we propose a new masking strategy which enhances the downstream performance of MPT and explains current improvements of masking strategies with our theoretical perspective." (JMLR's own per-paper page was blocked; retrieved from the paper's arXiv mirror, arXiv:2609.06460.) *Publication/online date: arXiv preprint dated 6 September 2026 (JMLR's own Published date unknown/inaccessible)*

JMLR Other · 发表于 2026

A Symplectic Analysis of Alternating Mirror Descent

Jonas E. Katona, Xiuyuan Wang, Andre Wibisono

摘要

Motivated by understanding the behavior of the Alternating Mirror Descent (AMD) algorithm for bilinear zero-sum games, we study the discretization of continuous-time Hamiltonian flow via the symplectic Euler method. We provide a framework for analysis using results from Hamiltonian dynamics and symplectic numerical integrators, with an emphasis on the existence and properties of a conserved quantity, the modified Hamiltonian (MH), for the symplectic Euler method. We compute the MH in closed-form when the original Hamiltonian is a quadratic function, and show that it generally differs from the other conserved quantity known previously in the literature. We derive new error bounds on the MH when truncated at orders in the stepsize in terms of the number of iterations, $K$, and use these bounds to show an improved $\mathcal{O}(K^{1/5})$ total regret bound and an $\mathcal{O}(K^{-4/5})$ duality gap of the average iterates for AMD. Finally, we propose a conjecture which, if true, would imply that the total regret for AMD scales as $\mathcal{O}\left(K^{\varepsilon}\right)$ and the duality gap of the average iterates as $\mathcal{O}\left(K^{-1+\varepsilon}\right)$ for any $\varepsilon>0$, and we can take $\varepsilon=0$ upon certain convergence conditions for the MH.

JMLR Optimization & Learning Theory · 发表于 2026

A Single-Loop Stochastic Proximal Quasi-Newton Method for Large-Scale Nonsmooth Convex Optimization

Yongcun Song, Zimeng Wang, Xiaoming Yuan, Hangrui Yue

摘要

We propose a new stochastic proximal quasi-Newton method for minimizing the sum of two convex functions in the particular context that one of the functions is the average of a large number of smooth functions and the other one is nonsmooth. The new method integrates a simple single-loop SVRG (L-SVRG) technique for sampling the gradient and a stochastic limited-memory BFGS (L-BFGS) scheme for approximating the Hessian of the smooth function components. The globally linear convergence rate of the new method is proved under mild assumptions. It is also shown that the new method covers a proximal variant of the L-SVRG as a special case, and it allows for various generalization through the integration with other variance reduction methods. For example, the L-SVRG can be replaced with the SAGA or SEGA in the proposed new method and thus other new stochastic proximal quasi-Newton methods with rigorously guaranteed convergence can be proposed accordingly. Moreover, we meticulously analyze the resulting nonsmooth subproblem at each iteration and leverage a compact representation of the L-BFGS matrix with the storage of some auxiliary matrices. As a result, we propose a very efficient and easily implementable semismooth Newton solver for solving the involved subproblems, whose arithmetic operations per iteration are merely order of O(d), where d denotes the dimensionality of the problem. With this efficient inner solver, the new method performs well and its numerical efficiency is validated through extensive experiments on a regularized logistic regression problem.

JMLR Econometrics & Economics Machine Learning Optimization & Learning Theory · 发表于 2026

A Reinforcement Learning Approach in Multi-Phase Second-Price Auction Design

Rui Ai, Boxiang Lyu, Zhaoran Wang, Zhuoran Yang, Michael I. Jordan

摘要

We study reserve price optimization in multi-phase second price auctions, where the seller's prior actions affect the bidders' later valuations through a Markov Decision Process (MDP). Compared to the bandit setting in existing works, the setting in ours involves three challenges. First, from the seller's perspective, we need to efficiently explore the environment in the presence of potentially untruthful bidders who aim to manipulate the seller's policy. Second, we want to minimize the seller's revenue regret when the market noise distribution is unknown. Third, the seller's per-step revenue is an unknown, nonlinear random variable, and cannot even be directly observed from the environment but realized values. We propose a mechanism addressing all three challenges. To address the first challenge, we use a combination of a new technique named “buffer periods” and inspirations from Reinforcement Learning (RL) with low switching cost to limit bidders' surplus from untruthful bidding, thereby incentivizing approximately truthful bidding. The second one is tackled by a novel algorithm that removes the need for pure exploration when the market noise distribution is unknown. The third challenge is resolved by an extension of LSVI-UCB, where we use the auction's underlying structure to control the uncertainty of the revenue function. The three techniques culminate in the \underline{C}ontextual-\underline{L}SVI-\underline{U}CB-\underline{B}uffer (CLUB) algorithm which achieves $\tilde{\mathcal{O}}(H^{5/2}\sqrt{K})$ revenue regret, where $K$ is the number of episodes and $H$ is the length of each episode, when the market noise is known and $\tilde{\mathcal{O}}(H^{3}\sqrt{K})$ revenue regret when the noise is unknown with no assumptions on bidders' truthfulness.

SiM Clinical Trials Optimization & Learning Theory Bayesian Methods · 发表于 2026,收录于 2026-08-17

A Randomized Basket Trial Design for Dose Optimization Based on Bayesian Model Averaging Using Spike-and-Slab Priors

Belay Birlie Yimer, Kentaro Takeda

摘要

The FDA initiated Project Optimus and issued guidance for dose optimization, recommending randomized parallel dose‐response cohorts to generate additional data at promising dose levels and implying that different dosages may be needed for different indications. In addition to dose optimization, with recent advancements in precision medicine and cancer biology, the development of cancer treatments has shifted toward the search for agents targeted to specific molecular profiles that may appear in more than one type of cancer. The basket trials are clinical trial designs that enable the simultaneous assessment of a new treatment in multiple indications. Concerning the FDA's dose optimization perspective and the recent trend of basket trials in early‐phase clinical trials, this paper proposes a dose‐ranging basket trial design based on a Bayesian model‐averaging approach considering efficacy and toxicity outcomes, where indications and dose levels define baskets. A key benefit of the proposed approach is that it explicitly accounts for the possible heterogeneity of response rates among baskets. Our simulation study shows that the proposed approach outperforms other methods, offering higher statistical power, better control of Type I error rates, precise optimal dose selection, and sample size savings in various scenarios with heterogeneous treatment effects between baskets.

JMLR Optimization & Learning Theory · 发表于 2026

A Provably Convergent Plug-and-Play Framework for Stochastic Bilevel Optimization

Tianshu Chu, Dachuan Xu, Wei Yao, Chengming Yu, Jin Zhang

摘要

Bilevel optimization has recently attracted significant attention in machine learning due to its wide range of applications and advanced hierarchical optimization capabilities. In this paper, we propose a plug-and-play framework, named PnPBO, for developing and analyzing stochastic bilevel optimization methods. This framework integrates both modern unbiased and biased stochastic estimators into the single-loop bilevel optimization framework introduced in Dagréou et al. (2022), with several improvements. In the implementation of PnPBO, all stochastic estimators for different variables can be independently incorporated, and an additional moving average technique is applied when using an unbiased estimator for the upper-level variable. In the theoretical analysis, we provide a unified convergence and complexity analysis for PnPBO, demonstrating that the adaptation of various stochastic estimators (including PAGE, ZeroSARAH, and mixed strategies) within the PnPBO framework achieves optimal sample complexity. Specifically, in the finite-sum setting, the resulting complexity matches the lower bound in Dagréou et al. (2024) and is comparable to that of single-level optimization (Zhou and Gu, 2019). This resolves the open question of whether the optimal complexity bounds for solving bilevel optimization are identical to those for single-level optimization. Finally, we empirically validate our framework, demonstrating its effectiveness on several benchmark problems and confirming our theoretical findings.

JMLR Optimization & Learning Theory · 发表于 2026,收录于 2026-11-06

A New Family of Poisson Non-negative Matrix Factorization Methods Using the Shifted Log Link

Eric Weine, Peter Carbonetto, Rafael A. Irizarry, Matthew Stephens

摘要

Poisson non-negative matrix factorization (NMF) is a widely used method to find interpretable "parts-based" decompositions of count data. While many variants of Poisson NMF exist, they all assume that the "parts" in the decomposition combine additively. This assumption may be natural in some settings, but not in others. Here, we introduce Poisson NMF with the shifted-log link function to relax this assumption. The shifted-log link function has a single tuning parameter, and as this parameter varies the model changes from assuming that parts combine additively to assuming that parts combine more multiplicatively. We provide an algorithm to fit this model by maximum likelihood, and also an approximation that substantially reduces computation time for large, sparse datasets. We illustrate these new methods on a variety of real and simulated data. Our examples show that while standard Poisson NMF often leads to sparse, cluster-like representations of data, the shifted-log link can produce more modular, layered results.

JMLR Networks & Graphs Machine Learning · 发表于 2026

A Neural Network Approach to Learning Solutions of a Class of Elliptic Variational Inequalities

Amal Alphonse, Michael Hintermüller, Alexander Kister, Chin Hang Lun, Clemens Sirotenko

摘要

We develop a weak adversarial approach to solving obstacle problems using neural networks. By employing (generalised) regularised gap functions and their properties we rewrite the obstacle problem (which is an elliptic variational inequality) as a minmax problem, providing a natural formulation amenable to learning. Our approach, in contrast to much of the literature, does not require the elliptic operator to be symmetric. We provide an error analysis for suitable discretisations of the continuous problem, estimating in particular the approximation and statistical errors. Parametrising the solution and test function as neural networks, we apply a modified gradient descent ascent algorithm to treat the problem and conclude the paper with various examples and experiments. Our solution algorithm is in particular able to easily handle obstacle problems that feature biactivity (or lack of strict complementarity), a situation that poses difficulty for traditional numerical methods.

JMLR Causal Inference Latent Variables & Mixtures · 发表于 2026,收录于 2026-11-06

A Negative-Control-Outcome Perspective on the Synthetic Control Method

Yixin Wang, Aaron Schein, Joshua Shou, David M. Blei

摘要

The synthetic controls method is a popular approach for analyzing panel data to estimate the causal effect of major policy interventions. The method is known to be justified when the data-generating process can be described by a linear factor model. This paper provides an alternative justification for the synthetic control method, which leads to a generalization of the algorithm that can incorporate non-linear factor models. Our justification centers around negative control outcomes, outcome variables for which the intervention is known a priori to have no effect. In synthetic controls, both pre-intervention outcomes of the target and pre- and post-intervention outcomes of the donors are all negative controls. Our key observation is that causal effects are identifiable under two assumptions: (1) the latent factors of the true (not necessarily linear) factor model satisfy weak unconfoundedness, and (2) the latent factors can be theoretically determined by the negative control outcomes. For example, these assumptions can be approximately satisfied in panel data with a long pre-treatment period. Based on these assumptions, we then generalize the synthetic controls method to an algorithm that estimates causal effects in non-linear settings. Crucial to this procedure is the idea of model criticism: the practitioner has license to explore and select a probabilistic factor model from a large class, provided it passes its goodness-of-fit test. We demonstrate this approach on a semi-synthetic study and use it to re-analyze a recent study on the impact of opening football stadiums during the COVID-19 pandemic. We find that this generalized approach to synthetic controls provides better counterfactual prediction than classical synthetic controls.

JMLR Machine Learning Networks & Graphs · 发表于 2026

A Natural Primal-Dual Hybrid Gradient Method for Adversarial Neural Network Training on Solving Partial Differential Equation

Shu Liu, Stanley Osher, Wuchen Li

摘要

We propose a scalable preconditioned primal-dual hybrid gradient algorithm for solving partial differential equations (PDEs). We multiply the PDE with a dual test function to obtain an inf-sup problem whose loss functional involves lower-order differential operators. The Primal-Dual Hybrid Gradient (PDHG) algorithm is then leveraged for this saddle point problem. By introducing suitable precondition operators to the proximal steps in the PDHG algorithm, we obtain an alternative natural gradient ascent-descent optimization scheme for updating the neural network parameters. We apply the Krylov subspace method (MINRES) to evaluate the natural gradients efficiently. Such treatment readily handles the inversion of precondition matrices via matrix-vector multiplication. An a posteriori convergence analysis is established for the time-continuous version of the proposed algorithm for general linear PDEs. By incorporating appropriate boundary loss terms, we further obtain a refined a priori convergence result for elliptic equations in divergence form. The algorithm is tested on various types of PDEs with dimensions ranging from $1$ to $50$, including linear and nonlinear elliptic equations, reaction-diffusion equations, and Monge-Ampere equations stemming from the $L^2$ optimal transport problems. We compare the performance of the proposed method with several commonly used deep learning algorithms such as physics-informed neural networks (PINNs), the DeepRitz method and weak adversarial networks (WANs) using either the Adam or the L-BFGS optimizer. The numerical results suggest that the proposed method performs efficiently and robustly and converges more stably with higher accuracy.

JMLR Optimization & Learning Theory Machine Learning Networks & Graphs · 发表于 2026

A Mean-Field Analysis of Neural Stochastic Gradient Descent-Ascent for Functional Minimax Optimization

Yuchen Zhu, Yufeng Zhang, Zhaoran Wang, Zhuoran Yang, Xiaohong Chen

摘要

This paper studies minimax optimization problems defined over infinite-dimensional function classes of over-parameterized two-layer neural networks. In particular, we consider the minimax optimization problem stemming from estimating linear functional equations defined by conditional expectations, where the objective functions are quadratic in the functional spaces. We address (i) the convergence of the stochastic gradient descent-ascent algorithm and (ii) the representation learning of the neural networks. We establish convergence in the mean-field regime by considering the continuous-time, infinite-width limit of the optimization dynamics. Under this regime, stochastic gradient descent-ascent corresponds to a Wasserstein gradient flow over the space of probability measures defined over the space of neural network parameters. We prove that the Wasserstein gradient flow converges globally to a stationary point of the minimax objective at a $\mathcal{O}(T^{-1} + \alpha^{-1} ) $ sublinear rate, and additionally finds the solution to the functional equation when the regularizer of the minimax objective is strongly convex. Here $T$ denotes the time and $\alpha$ is a scaling parameter of the neural networks. In terms of representation learning, our results show that the feature representation induced by the neural networks may deviate from the initial representation by a factor of $\mathcal{O}(\alpha^{-1})$, measured by the Wasserstein distance. Finally, we apply our general results to concrete examples, including policy evaluation, nonparametric instrumental variable regression, and asset pricing.

JMLR Other · 发表于 2026,收录于 2026-09-17

A Library for Learning Neural Operators

Jean Kossaifi, Nikola Kovachki, Zongyi Li, David Pitt, Miguel Liu-Schiaffini, Robert Joseph George, Boris Bonev, Kamyar Azizzadenesheli, Julius Berner, Valentin Duruisseaux, Anima Anandkumar

摘要

We present NeuralOperator, an open-source Python library for operator learning. Neural operators generalize neural networks to maps between function spaces instead of finite-dimensional Euclidean spaces. They can be trained and inferenced on input and output functions given at various discretizations, satisfying a discretization convergence properties. Part of the official PyTorch Ecosystem, NeuralOperator provides all the tools for training and deploying neural operator models, as well as developing new ones, in a high-quality, tested, open-source package. It combines cutting-edge models and customizability with a gentle learning curve and simple user interface for newcomers." (JMLR's own per-paper page was blocked; retrieved from the paper's arXiv mirror, arXiv:2412.10354.) *Publication/online date: date unknown (JMLR page inaccessible; listed in Volume 27, 2026)*

JMLR Machine Learning Networks & Graphs · 发表于 2026

A Functional-Space Mean-Field Theory of Partially-Trained Three-Layer Neural Networks

Zhengdao Chen, Eric Vanden-Eijnden, Joan Bruna

摘要

To understand the training dynamics of neural networks, prior studies have considered the mean-field (MF) limit of two-layer NNs as the width tends to infinity, establishing theoretical guarantees for its convergence under gradient flow training as well as approximation and generalization capabilities. In this work, we study the infinite-width limit of a type of three-layer neural network where the first-layer weights are untrained. To rigorously define the limiting model, we extend the MF theory by lifting the representation of neurons from Euclidean to functional spaces. This allows us to establish the MF training dynamics as a functional gradient flow with a time-varying kernel that remains positive-definite under suitable assumptions, thus proving a linear-rate convergence of its training loss. Furthermore, we define novel function spaces that contain the solutions obtained through the MF training dynamics and prove Rademacher complexity bounds for these spaces. Notably, our analysis applies to a range of scaling choices of the model, resulting in two distinct regimes of the MF limit that both exhibit feature learning through training.

JMLR Other · 发表于 2026

A Fully Parameter-Free Second-Order Algorithm for Convex-Concave Minimax Problems

Jun-Lin Wang, Zi Xu, Hui-Ling Zhang

摘要

In this paper, we study second-order algorithms for the convex-concave minimax problem, which has attracted much attention in many fields such as machine learning in recent years. We propose a Lipschitz-free cubic regularization (LF-CR) algorithm for solving the convex-concave minimax optimization problem without knowing the Lipschitz constant. It can be shown that the iteration complexity of the LF-CR algorithm to obtain an $\epsilon$-optimal solution with respect to the restricted primal-dual gap is upper bounded by $\mathcal{O}(\rho^{2/3}\|z_0-z^*\|^2\epsilon^{-2/3})$ , where $z_0=(x_0,y_0)$ is a pair of initial points, $z^*=(x^*,y^*)$ is a pair of optimal solutions, and $\rho$ is the Lipschitz constant. We further propose a fully parameter-free cubic regularization (FF-CR) algorithm that does not require any parameters of the problem, including the Lipschitz constant and the upper bound of the distance from the initial point to the optimal solution. We also prove that the iteration complexity of the FF-CR algorithm to obtain an $\epsilon$-optimal solution with respect to the gradient norm is upper bounded by $\mathcal{O}(\rho^{2/3}\|z_0-z^*\|^{4/3}\epsilon^{-2/3}) $. Numerical experiments show the efficiency of both algorithms. To the best of our knowledge, the proposed FF-CR algorithm is a completely parameter-free second-order algorithm, and its iteration complexity is currently the best in terms of $\epsilon$ under the termination criterion of the gradient norm.

Econometrica Econometrics & Economics Genetics & Omics · 发表于 2026

A Framework for Geoeconomics

Christopher Clayton, Matteo Maggiori, Jesse Schreger

摘要

Governments use their countries' economic strength from financial and trade relationships to achieve geopolitical and economic goals. We provide a model of the sources of geoeconomic power and how it is wielded. The source of this power is the ability of a hegemonic country to coordinate threats across disparate economic relationships as a means of enforcement on foreign entities. The hegemon wields this power to demand costly actions out of the targeted entities, including mark‐ups, import restrictions, tariffs, and political concessions. The hegemon uses its power to change targeted entities' activities to manipulate the global equilibrium in its favor and increase its power. A sector is strategic either in helping the hegemon form threats or in manipulating the world equilibrium via input‐output amplification. The hegemon acts a global enforcer, thus adding value to the world economy, but destroys value by distorting the equilibrium in its favor.

JMLR Nonparametric & Semiparametric Networks & Graphs · 发表于 2026

A Data-Augmented Contrastive Learning Approach to Nonparametric Density Estimation

Chenghao Li, Yuanyuan Lin

摘要

In this paper, we introduce a data-augmented nonparametric noise contrastive estimation method to density estimation using deep neural networks. By leveraging the idea of contrastive learning, our density estimator exhibits efficiency with a one-step and simulation-free evaluation process, imposes no constraints on the neural network, and is shown to be consistent and asymptotically automatically normalized. A novel data augmentation procedure allows us to mitigate the influence of the choice of reference distribution on our method. Non-asymptotic upper bounds for the expected $L_{2}$-risk and the expected total variation distance have been established, which achieve minimax optimal rates. Moreover, our new method exhibits inherent adaptivity to low dimensional structures of data with a faster convergence rate under a compositional structure assumption. Numerical experiments show the competitiveness of our new method compared with the state-of-the-art nonparametric density estimation methods.

JMLR Causal Inference · 发表于 2026

A Convex Framework for Confounding Robust Inference

Kei Ishikawa, Niao He, Takafumi Kanamori

摘要

We study policy evaluation of offline contextual bandits subject to unobserved confounders. Sensitivity analysis methods are commonly used to estimate the policy value under the worst-case confounding scenario within a given uncertainty set. However, existing work often resorts to some coarse relaxation of the uncertainty set for the sake of tractability, leading to overly conservative estimation of the policy value. In this paper, we propose a general estimator that provides a sharp lower bound of the policy value using convex programming. The generality of our estimator enables various extensions such as sensitivity analysis using f-divergence, model selection with cross validation and information criterion, and robust policy learning with the sharp lower bound. Furthermore, our estimation method can be reformulated as an empirical risk minimization problem thanks to the strong duality, which enables us to provide strong theoretical guarantees of the proposed estimator using M-estimation techniques.

JMLR Other · 发表于 2026

A Common Interface for Automatic Differentiation

Guillaume Dalle, Adrian Hill

摘要

For scientific machine learning tasks with a lot of custom code, picking the right Automatic Differentiation (AD) system matters. Our Julia package DifferentiationInterface.jl provides a common frontend to a dozen AD backends, unlocking easy comparison and modular development. In particular, its built-in preparation mechanism leverages the strengths of each backend by amortizing one-time computations. This is key to enabling sophisticated features like sparsity handling without putting additional burdens on the user.

第 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 页