Search

Randomized matrix algorithms have become workhorse tools in scientific computing and machine learning. To use these algorithms safely in applications, they should be coupled with posterior error estimates to assess the quality of the output. To meet this need, this paper proposes two diagnostics: a leave-one-out error estimator for randomized low-rank approximations and a jackknife resampling method to estimate the variance of the output of a randomized matrix computation. Both of these diagnostics are rapid to compute for randomized low-rank approximation algorithms such as the randomized SVD and randomized Nystr\"om approximation, and they provide useful information that can be used to assess the quality of the computed output and guide algorithmic parameter choices. [ABSTRACT FROM AUTHOR]

The dynamic formulation of optimal transport has attracted growing interest in scientific computing and machine learning, and its computation requires solving a PDE-constrained optimization problem. The classical Eulerian discretization based approaches suffer from the curse of dimensionality, which arises from the approximation of a high-dimensional velocity field. In this work, we propose a deep learning based method for solving the dynamic optimal transport in highdimensional space. Our method contains three main ingredients: a carefully designed representation of the velocity field, the discretization of the PDE constraint along the characteristics, and the computation of a high-dimensional integral by the Monte Carlo method in each time step. Specifically, in the representation of the velocity field, we apply the classical nodal basis function in time and the deep neural networks in the space domain with the H1-norm regularization. This technique promotes the regularity of the velocity field in both time and space such that the discretization along the characteristic remains stable during the training process. Extensive numerical examples have been conducted to test the proposed method. Compared to other solvers of optimal transport, our method could give more accurate results in high-dimensional cases and has very good scalability with respect to dimension. Finally, we extend our method to more complicated cases such as the crowd motion problem. [ABSTRACT FROM AUTHOR]

The computation of f(A)b, the action of a matrix function on a vector, is a task arising in many areas of scientific computing. In many applications, the matrix A is sparse but so large that only a rather small number of Krylov basis vectors can be stored. Here we discuss a new approach to overcome this limitation by randomized sketching combined with an integral representation of f(A)b. Two different approximation methods are introduced, one based on sketched FOM and another based on sketched GMRES. The convergence of the latter method is analyzed for Stieltjes functions of positive real matrices. We also derive a closed-form expression for the sketched FOM approximant and bound its distance to the full FOM approximant. Numerical experiments demonstrate the potential of the presented sketching approaches. [ABSTRACT FROM AUTHOR]

We present the derivation, implementation, and analysis of a multiresolution adaptive grid framework for numerical simulations on octree-based three-dimensional block-structured collocated grids with distributed computational architectures. Our approach provides a consistent handling of nonlifted and lifted interpolating wavelets of arbitrary order demonstrated using second-, fourth-, and sixth-order wavelets, combined with standard finite-difference-based discretization op-erators. We first validate that the wavelet family used provides strict and explicit error control when coarsening the grid, and show that lifting wavelets increase the grid compression rate while conserving discrete moments across levels. Further, we demonstrate that high-order PDE discretization schemes combined with sufficiently high-order wavelets retain the expected convergence order even at resolution jumps. We then simulate the advection of a scalar to analyze convergence for the temporal evolution of a PDE. The results shows that our wavelet-based refinement criterion is successful at controlling the overall error while the coarsening criterion is effective at retaining the relevant information on a compressed grid. Our software exploits a block-structured grid data structure for efficient multilevel operations, combined with a parallelization strategy that relies on a one-sided MPI-RMA communication approach with active post-start-complete-wait synchronization. Using performance tests up to 16,384 cores, we demonstrate that this leads to a highly scalable performance. The associated code is available under a BSD-3 license at https://github.com/vanreeslab/murphy. [ABSTRACT FROM AUTHOR]

The article introduces a new section structure for the SIAM Journal on Scientific Computing (SISC), including a focus on Scientific Machine Learning as a major development in the field of scientific computing. Topics include the development of accurate, efficient, and scalable machine learning algorithms for solving mathematical modeling problems, emphasizing the importance of reproducibility in the review process.

The article offers information on The Seventeenth Copper Mountain Conference on Iterative Methods (CMCIM), which took place virtually in April 2022.

Least squares form one of the most prominent classes of optimization problems with numerous applications in scientific computing and data fitting. When such formulations aim at modeling complex systems, the optimization process must account for nonlinear dynamics by incorporating constraints. In addition, these systems often incorporate a large number of variables, which increases the difficulty of the problem and motivates the need for efficient algorithms amenable to large-scale implementations. In this paper, we propose and analyze a Levenberg--Marquardt algorithm for nonlinear least squares subject to nonlinear equality constraints. Our algorithm is based on inexact solves of linear least-squares problems that only require Jacobian-vector products. Global convergence is guaranteed by the combination of a composite step approach and a nonmonotone step acceptance rule. We illustrate the performance of our method on several test cases from data assimilation and inverse problems; our algorithm is able to reach the vicinity of a solution from an arbitrary starting point and can outperform the most natural alternatives for these classes of problems. [ABSTRACT FROM AUTHOR]

The article focuses on research reproducibility in scientific computing journals, Journal on Scientific Computing (SISC), Journal of Scientific Computing (JSC), and Journal of Computational Physics (JCP), by requiring authors to publicly share their data and credibility in scientific research.