225 papers
This paper proposes a data-driven framework using time-varying queueing models to improve long-term ICU capacity planning by capturing fluctuating admission rates and heterogeneous length-of-stay distributions, demonstrating that static heuristics like the 85% occupancy rule are inadequate for managing real-world demand variability in neonatal intensive care units.
This paper extends the Fractional HBVMs (FHBVMs) framework to efficiently solve systems of fractional differential equations with multiple, distinct derivative orders, addressing a previous limitation and providing a corresponding effective MATLAB implementation.
This paper proposes a memory-efficient multistep approach for computing Floquet multipliers and subspaces that converts the problem into a periodic polynomial eigenvalue problem, proving that parasitic eigenvalues vanish geometrically as stepsize decreases while demonstrating the efficiency of the new pTOAR algorithm for large-scale systems.
This paper proves that the discrete Dirichlet-to-Neumann operator uniquely determines edge conductivities on multi-dimensional hypercubic lattices (dimension three or higher) by employing a novel slicing technique, thereby extending the classical two-dimensional uniqueness result of Curtis and Morrow.
This paper introduces deep kernel greedy models that combine the computational efficiency and provable convergence of greedy sparse approximation with the enhanced expressiveness of deep, multilayer kernels, demonstrating superior approximation accuracy over standard kernels and neural networks across various scientific applications.
This paper introduces and analyzes a divide-and-conquer algorithm for quantizing one-dimensional probability distributions, establishing a universal Wasserstein-1 error bound and demonstrating through numerical experiments that the method achieves optimal convergence rates while offering superior stability under arithmetic operations compared to existing schemes.
This paper proposes and analyzes two scalable augmented Lagrangian preconditioners for solving block-structured linear systems arising from fictitious domain discretizations of Poisson and Stokes problems, demonstrating their robustness and effectiveness through spectral analysis and extensive numerical tests in two and three dimensions.
This paper presents a novel, scalable computational framework implemented in the open-source SPEED software that combines model order reduction with a high-fidelity Discontinuous Galerkin Spectral Element Method to efficiently simulate the coupled electromechanical-acoustic behavior of large-scale Piezoelectric Micromachined Ultrasonic Transducer (PMUT) arrays.
This paper presents a novel, general 3D finite element framework for railway vehicle-bridge interaction that utilizes absolute coordinates and tailored constraint equations to rigorously model nonlinear wheel-rail contact, enabling robust simulations of extreme lateral movements without relying on infinitesimal displacement assumptions.
This paper develops a spectral method to solve coupled kinetic and viscous half-space problems at network nodes, enabling the derivation of accurate macroscopic coupling conditions for the linearized BGK equation and its acoustic limit through detailed asymptotic analysis of interface layers.
This paper provides a comprehensive update on stochastic rounding research from 2022 to 2026, with a specific focus on the new limited-precision variant, while offering insights into its industrial applications and future hardware implementation.
This paper introduces two numerical frameworks—a strong-competition penalty method with damped Gauss-Seidel iterations and a projected gradient method with an accelerated variant—to solve elliptic systems governed by nonconvex partial segregation constraints where three nonnegative components must have a vanishing pointwise product.
The paper proposes PriorIDENT, a prior-informed weak-form sparse-regression framework that integrates compact physics priors into dictionary construction to robustly identify governing partial differential equations from noisy spatiotemporal data while outperforming existing baselines in accuracy and stability.
This paper introduces FlexTrace, a novel single-pass trace estimator that accurately computes the trace of matrix functions for large symmetric positive semi-definite matrices using only matrix-vector products with the original matrix, thereby overcoming the computational expense of existing methods that require products with the function of the matrix.
This paper demonstrates that combining a partial pivoted Cholesky approximation with a Vecchia approximation of the residual yields an exact Vecchia approximation of the original matrix with an augmented sparsity pattern, thereby unifying low-rank and sparse inverse Cholesky approximations under a single framework.
This paper introduces Drifting Models for molecular conformation generation, leveraging a novel Drifting Force Identity to incorporate physical force labels into one-step sampling, thereby achieving million-fold acceleration over traditional molecular dynamics while maintaining perfect structural validity and Boltzmann distribution accuracy.
This paper introduces the L-Transformer, a novel architecture that utilizes structured spectral factorization via the L-product to decompose the embedding space into independent spectral sub-transformers, achieving significant parameter reduction (up to 75%) while maintaining competitive performance and introducing beneficial frequency-based inductive biases.
This paper presents a framework for the certified and accurate computation of deep neural network function space norms (including , , and ) by combining interval arithmetic, adaptive refinement, and quadrature to derive guaranteed global bounds from local certificates, thereby enabling reliable error control for PDE applications like PINNs.
This paper introduces ALFCG, the first adaptive, projection-free framework for stochastic composite nonconvex optimization that eliminates the need for global smoothness constants or line search by using self-normalized accumulators to estimate local smoothness, achieving optimal iteration complexity up to logarithmic factors while outperforming state-of-the-art baselines.
This paper introduces FourierSpecNet, a hybrid deep learning framework that integrates the Fourier spectral method to efficiently approximate the Boltzmann collision operator, achieving resolution-invariant learning, zero-shot super-resolution, and significant computational savings while maintaining accuracy across elastic and inelastic collision regimes.