Unified framework estimates desirability of outcome ranking for benefit-risk evaluation.
problem Estimating desirability of outcome ranking in randomized and observational studies.
method Unified covariate-adjusted causal inference framework, estimating conditional ordinal distributions through sequential risk-set hazards, and deriving efficient influence function (EIF).
result CVTMLE-SL showed strongest performance across various settings.
FDR criterion simplifies complex causal graphs to a standard front-door setting.
problem Complex causal graphs make identification of causal effects difficult and computationally infeasible.
method Front-door reducibility (FDR) criterion and FDR-TID algorithm.
result Many graphs can be simplified to a standard front-door setting, making causal effect identification simpler and more interpretable.
New methods estimate causal effects using front-door criterion in presence of unmeasured confounders.
problem Estimating causal effects in observational studies with unmeasured confounders.
method Developed novel one-step and targeted minimum loss-based estimators for front-door assumptions.
result Established conditions for root-n consistency and asymptotic linearity.
New method estimates causal effects without knowing graph structure.
problem Estimating causal effects when graph structure is unknown.
method Testable conditional independence statements for front-door adjustment.
result Effect estimation without Markov equivalence class knowledge.
Flow Matching enables robust training of CNFs with various probability paths.
problem Training Continuous Normalizing Flows (CNFs) at large scales.
method Flow Matching (FM) is a simulation-free approach for training CNFs by regressing vector fields of conditional probability paths.
result Flow Matching with diffusion paths yields more robust and stable training compared to diffusion-based methods.
Extends causal inference to hidden mediators with proxies.
problem Identifying causal effects with hidden mediators and error-prone proxies.
method Established causal hidden mediation analysis and hidden front-door criterion.
result Identification of population intervention indirect effect possible with hidden mediators.
New estimators for causal effects in DAGs with hidden variables, addressing computational and statistical challenges.
problem Estimating causal effects in DAGs with hidden variables beyond traditional criteria.
method Introduces novel one-step corrected plug-in and targeted minimum loss-based estimators for causal effects in DAGs with hidden variables.
result Root-n consistent causal effect estimates with desirable statistical properties.
A new method estimates rare events using tensor trains.
problem Estimating rare event probabilities in high-dimensional problems.
method Approximating optimal importance distribution via tensor-train decompositions and compositions.
result Better variance reduction and efficient computation of rare event probabilities.
Debiased learners estimate heterogeneous treatment effects in observational studies.
problem Estimating heterogeneous treatment effects in observational studies with unmeasured confounders.
method Debiased Front-Door (FD) learners, FD-DR-Learner and FD-R-Learner, under specific assumptions.
result Debiased learners satisfy error bounds and stage-error decompositions, delivering reliable HTE estimates.
Develops a measure-theoretic framework for complex co-occurrence data.
problem Modeling and interpreting complex co-occurrences in high-dimensional data.
method Introduces measure-theoretic probability and conditional probability, investigates E-integrals.
result Establishes a rigorous measure-theoretic foundation for co-occurrence modeling.
Paper uses GMM and MAF for probabilistic classification, outperforming simpler models.
problem Classifying data with complex distributions.
method Density estimation using Gaussian Mixture Model and Masked Autoregressive Flow.
result Proposed classifiers outperform simpler models like linear discriminant analysis.
We win EVA2025 by estimating extreme precipitation events using Peaks Over Thresholds and martingale testing.
problem Estimating the probability of extreme precipitation events with limited data.
method Modeling Peaks Over Thresholds with an exponential distribution and using martingale testing for evaluation.
result Our method outperforms other approaches in estimating extreme precipitation events.
We propose and document the evidence for an analogy between the dynamics of granular counter-flows in the presence of bottlenecks or restrictions and financial price formation processes. Using extensive simulations, we find that the counter-flows of simulated pedestrians through a door display many stylized facts obser…
Memory constraints slow linear regression convergence.
problem Linear regression with limited memory.
method Analyzing memory and sample requirements for regression.
result Subquadratic memory algorithms converge slower than unrestricted methods.
A lemma of Tits establishes a connection between the simple connectivity of an incidence geometry and the universal completion of an amalgam induced by a sufficiently transitive group of automorphisms of that geometry. In the present paper, we generalize this lemma to intransitive geometries, thus opening the door for …
Secure linear regression at speed of plaintext methods.
problem Secure multiparty linear regression and feature selection.
method Distributed algorithms combining geometric ideas.
result Efficient and secure genome-wide association studies.
We introduce a model in which a regulator employs mechanism design to embed her human capital beta signal(s) in a firm's capital structure, in order to enhance the value of her post career change indexed executive stock option contract with the firm. We prove that the agency cost of this revolving door behavior increas…
Introduces a framework using information theory for understanding machine learning.
problem Understanding the effectiveness and design of modern machine learning architectures.
method An information-theoretic approach to learning, focusing on model complexity and architecture.
result Successful architectures have a broad complexity range, enabling learning in over-parameterized model classes.
A new causal graph framework identifies treatment effects without adjusting for confounders.
problem Invalid identification of causal effects due to unmeasured confounders.
method Developed the Napkin graph to identify causal effects through a ratio of g-formulas, using influence-function-based estimators.
result Demonstrated substantial efficiency gains in estimating causal effects using the Napkin graph.
OCEAN infers online task identities from context variables.
problem Online task inference for compositional tasks with context adaptation.
method Variational inference framework OCEAN models global and local context variables in a joint latent space.
result OCEAN provides more effective task inference with sequential context adaptation.
We introduce a probabilistic approach to the LMS filter. By means of an efficient approximation, this approach provides an adaptable step-size LMS algorithm together with a measure of uncertainty about the estimation. In addition, the proposed approximation preserves the linear complexity of the standard LMS. Numerical…
Heavy-tailed outliers are more resilient to robust estimation than adversarial ones.
problem Developing robust estimators for data with outliers.
method Analyzing the relationship between adversarial and heavy-tailed outlier models.
result Optimal estimators for heavy-tailed outliers are also optimal for adversarial settings, but not vice versa.
Forecasting COVID-19 cases in Senegal using machine learning.
problem Predicting the inflection point and ending time of COVID-19 cases in Senegal.
method Visualization and machine learning techniques applied to public data.
result Forecasted the inflection point and possible ending time of COVID-19 cases in Senegal.
Unified framework for estimating indirect effects in observational studies with unmeasured confounding.
problem Challenges in evaluating indirect effects due to unmeasured confounding and unethical exposures.
method Developed a unified identification and estimation framework using proximal causal inference.
result Unified identification and estimation of PIIE and causal effect of an intervening variable in settings with pervasive unmeasured confounding.
The goal of this paper is to exhibit a deep relation between the partition function of the Ising model on a planar trivalent graph and the generating series of the spin network evaluations on the same graph. We provide respectively a fermionic and a bosonic Gaussian integral formulation for each of these functions and …
New algorithm improves active learning in agnostic pool-based classification.
problem Efficient active learning in the agnostic setting with minimized sample complexity.
method Solves an experimental design problem to determine a distribution over examples for label requests.
result Achieves sample complexity bounds never worse than best disagreement coefficient-based bounds, sometimes significantly smaller.
Deep learning classifies over 94% of crystallization images accurately.
problem Classifying macromolecular crystallization outcomes from various experiments.
method Deep convolutional neural networks trained on a large annotated dataset.
result More than 94% of test images correctly labeled, regardless of origin.
A new model uses low-rank DPPs to improve product recommendation.
problem Scalability issues in DPP models for large datasets.
method Low-rank DPP mixture model with MCMC learning.
result Substantially better predictive performance than single DPP models.
New method verifies formulas for causal interventional distributions.
problem Deciding if a given formula correctly identifies an interventional distribution.
method Proposed a falsifier to check if a formula is identifying.
result Falsifier can induce an almost-surely correct verifier for certain models.
New SDA models for big data analysis using aggregated symbols.
problem Handling large and complex datasets efficiently.
method Developing likelihood functions for symbolic data based on underlying measurement-level data.
result Efficient analysis of big data through reduced distributional summaries.
Introduces a new Heston model with multiple factors.
problem Reconciling classical Heston model with rough Heston model.
method Develops a lifted Heston model with n multi-factors.
result The lifted model provides better fits and faster calibration.
Quantum computing tackles non-convex portfolio optimization with cardinality constraints.
problem Non-convex portfolio optimization problems in asset management.
method Application of quantum annealing with non-linear cardinality constraints.
result Quantum portfolio optimization yields smaller, more profitable portfolios.
Word2vec analysis reveals spectral underpinnings.
problem Lack of theoretical justification for word2vec.
method Rigorous spectral analysis of word2vec's nonlinear functional.
result Word2vec may be primarily driven by spectral method.
New kernel methods estimate complex causal relationships.
problem Estimating nonparametric causal functions like dose-response curves.
method Kernel ridge regression with decomposition property.
result Uniform consistency with finite sample rates proved.
Deep models generate and optimize DNA sequences for protein binding.
problem Designing DNA sequences with desired properties.
method Three approaches: GAN for synthetic sequences, activation maximization for design, and a combined method.
result Generated DNA sequences have superior properties to those in training data.
New method uses Multiple Choice Learning for speech separation.
problem Ambiguous task of assigning model predictions to ground truth signals.
method Uses Multiple Choice Learning (MCL) instead of Permutation Invariant Training (PIT).
result MCL matches PIT performance but is computationally advantageous.
Paper develops methods to fairly measure contributions in federated learning.
problem Fairly allocate credits for participants in federated learning.
method Uses deletion method for horizontal FML and Shapley Values for vertical FML.
result Developed techniques to calculate contributions in federated learning.
Random matrix theory analyzes learning dynamics in neural networks.
problem Understanding learning dynamics in neural networks.
method Random matrix approach to analyze a linear network.
result Insights into overfitting, early stopping, and initialization.
The paper develops algorithms to minimize risk and regret in uncertain decisions.
problem Minimizing risk and regret in multistage decisions under uncertainty.
method Established dual representations and used Lagrangian duality theory to develop progressive hedging algorithms.
result Modified progressive hedging algorithm can handle new linkage constraints.
The paper analyzes financial market turbulence using mathematical physics.
problem Understanding price fluctuations caused by information asymmetry.
method Spectrum analysis to decompose pricing patterns.
result Identifies phase correlations in financial stock market turbulence.
Bayesian approach improves semi-supervised learning with deep generative models.
problem Lack of model uncertainty and flexibility in existing semi-supervised learning methods.
method Proposes a discriminative component with stochastic inputs and extends it to be fully Bayesian.
result Improved handling of model uncertainty and flexibility in capturing complex patterns.
Large particle systems' fluctuations converge to SPDE with additive noise.
problem Understanding large equity markets and investment strategies.
method Hydrodynamic limit and SPDE analysis of rank-based models.
result Fluctuations of empirical cumulative distribution functions converge to SPDE.
Proposes a deep hedging method for robust pricing and hedging under parameter uncertainty.
problem Pricing and hedging under parameter uncertainty for generalized affine processes.
method Deep learning approach linked to variational form of Kolmogorov equation.
result Robust deep hedging outperforms existing methods in volatile periods.
Noise decreases the Hessian spectrum in overparameterized networks, aiding generalization.
problem Understanding why SGD leads to good generalization in overparameterized neural networks.
method Analyzing the Hessian spectrum under noise and other conditions.
result Noise decreases the trace and determinant of the Hessian spectrum in overparameterized networks.
UR-FUNNY dataset aids in understanding multimodal humor.
problem Understanding humor in a multimodal context is understudied.
method Developed a multimodal dataset (UR-FUNNY) for humor detection.
result UR-FUNNY opens the door to multimodal humor detection research.
New method uses Fisher-Rao metric for non-Gaussian decoders.
problem Existing latent space geometry theory only works for Gaussian decoders.
method Pull back Fisher-Rao metric to latent space for non-Gaussian decoders.
result Achieves meaningful latent geometries for various non-Gaussian decoders.
Two new methods estimate quantum density matrices using machine learning.
problem Estimating the quantum density matrix for complex systems.
method Quantum Maximum Likelihood and Quantum Variational Inference with quantum flows.
result Improved estimation of quantum density matrices for mixed states.
TopRank algorithm improves online ranking with better performance and insights.
problem Sequential decision-making in online learning to rank with user feedback.
method Generalized click model and topological sort-based algorithm.
result TopRank outperforms existing algorithms in terms of performance and proof insight.