arXiv research

This work compares regularization and constrained inference for label constraints in machine learning.

Tikhonov regularization is robust under specific martingale constraints in distributionally robust optimization.

Solves Einstein constraint equations on compact manifolds with specified boundaries.

The paper shows regularization can't always find all optimal solutions in constrained ML.

The paper tackles safe reinforcement learning with convex regularization.

DONUT improves treatment effect estimation by enforcing orthogonality constraints.

Paper proposes a new sparse group k-max regularization for sparsity constraints.

We construct low regularity solutions of the vacuum Einstein constraint equations. In particular, on 3-manifolds we obtain solutions with metrics in $H^s\loc$ with $s>{3\over 2}$. The theory of maximal asymptotically Euclidean solutions of the constraint equations descends completely the low regularity setting. Moreove…

Paper optimizes ES estimation under an $\ell_1$ constraint, reducing estimation errors.

This paper studies nonholonomic constraints in Hamiltonian systems, deriving equations and theorems.

The paper studies the properties of maps with free boundaries, focusing on the obstacle case.

Structured regularizers enable faster optimization on SPD manifolds with constraints.

ARO overfits by making constraints dependent on uncertainty, leading to brittleness.

Study uses property elicitation to understand how fairness regularizers affect optimal decisions.

Proposes methods to add constraints to neural networks to improve stability and generalization.

The paper examines how ESG constraints affect portfolio optimization in large datasets.

Paper proposes a new method to optimize deep neural networks with sparse regularization.

Proper regularization is critical for speeding up training, improving generalization performance, and learning compact models that are cost efficient. We propose and analyze regularized gradient descent algorithms for learning shallow neural networks. Our framework is general and covers weight-sharing (convolutional ne…

Since their invention, generative adversarial networks (GANs) have become a popular approach for learning to model a distribution of real (unlabeled) data. Convergence problems during training are overcome by Wasserstein GANs which minimize the distance between the model and the empirical distribution in terms of a dif…

New approach to sparse optimal transport for matching tokens with experts.

Optimal bounds on regret and constraint violation in adversarial COCO.

This paper considers distributed online optimization with time-varying coupled inequality constraints. The global objective function is composed of local convex cost and regularization functions and the coupled constraint function is the sum of local convex functions. A distributed online primal-dual dynamic mirror des…

Lockout solves sparse regularization for neural networks.

Selective state-adaptive regularization improves offline RL performance.

Kernel-based tests for shape constraints in finance.

Survey of Gaussian process constraints for modeling expensive data.

MARL algorithm uses regularization to avoid explicit structures, improving performance.

Adapting Hedge algorithm for semi-adversarial data with root-entropy regularization.

Fair Mixup improves fairness in classifiers by interpolating between groups.

Study optimal consumption with relaxed benchmarks and drawdown constraints.

Unified approach adjusts classifiers to meet system-level constraints.

Optimal transport framework for density estimation with constraints.

We establish the regularity theory for certain critical elliptic systems with an anti-symmetric structure under inhomogeneous Neumann and Dirichlet boundary constraints. As applications, we prove full regularity and smooth estimates at the free boundary for weakly Dirac-harmonic maps from spin Riemann surfaces. Our met…

Muon optimizer improves deep learning with spectral norm constraints.

New framework for robust regularization under uncertain data distributions.

Regularized least-squares approaches have been successfully applied to linear system identification. Recent approaches use quadratic penalty terms on the unknown impulse response defined by stable spline kernels, which control model space complexity by leveraging regularity and bounded-input bounded-output stability. T…

This paper proposes robust matrix variate regression models with rank constraints and vector regularization.

New framework improves reliability of learned representations by modeling uncertainty and structural constraints.

A scalable framework optimizes multi-asset portfolios with constraints.

We present an objective function for learning with unlabeled data that utilizes auxiliary expectation constraints. We optimize this objective function using a procedure that alternates between information and moment projections. Our method provides an alternate interpretation of the posterior regularization framework (…

Framework for designing nonlinearities in neural networks with slope constraints.

Compositional data have two unique characteristics compared to typical multivariate data: the observed values are nonnegative and their summand is exactly one. To reflect these characteristics, a specific regularized regression model with linear constraints is commonly used. However, linear constraints incur additional…

We consider the problem of supervised learning with convex loss functions and propose a new form of iterative regularization based on the subgradient method. Unlike other regularization approaches, in iterative regularization no constraint or penalization is considered, and generalization is achieved by (early) stoppin…

Classifies tilings of hyperbolic plane by regular polygons.

We consider adaptive system identification problems with convex constraints and propose a family of regularized Least-Mean-Square (LMS) algorithms. We show that with a properly selected regularization parameter the regularized LMS provably dominates its conventional counterpart in terms of mean square deviations. We es…

We propose Absum, which is a regularization method for improving adversarial robustness of convolutional neural networks (CNNs). Although CNNs can accurately recognize images, recent studies have shown that the convolution operations in CNNs commonly have structural sensitivity to specific noise composed of Fourier bas…

The problem of low-rank approximation with convex constraints, which appears in data analysis, system identification, model order reduction, low-order controller design and low-complexity modelling is considered. Given a matrix, the objective is to find a low-rank approximation that meets rank and convex constraints, w…

Estimation in generalized linear models (GLM) is complicated by the presence of constraints. One can handle constraints by maximizing a penalized log-likelihood. Penalties such as the lasso are effective in high dimensions, but often lead to unwanted shrinkage. This paper explores instead penalizing the squared distanc…