Numerical Analysis

New submissions

• New submissions
• Cross-lists
• Replacements

New submissions for Fri, 1 Jul 22

[1]  arXiv:2206.14905 [pdf, ps, other]

Title: Generalized Pseudoskeleton Decompositions

Authors: Keaton Hamm

Subjects: Numerical Analysis (math.NA)

We characterize some variations of pseudoskeleton (also called CUR) decompositions for matrices and tensors over arbitrary fields. These characterizations extend previous results to arbitrary fields and to decompositions which use generalized inverses of the constituent matrices, in contrast to Moore--Penrose pseudoinverses in prior works which are specific to real or complex valued matrices, and are significantly more structured.

[2]  arXiv:2206.15004 [pdf, other]

Title: Padé-parametric FEM approximation for fractional powers of elliptic operators on manifolds

Authors: Beiping Duan

Subjects: Numerical Analysis (math.NA)

This paper focuses on numerical approximation for fractional powers of elliptic operators on $2$-d manifolds. Firstly, parametric finite element method is employed to discretize the original problem. We then approximate fractional powers of the discrete elliptic operator by the product of rational functions, each of which is a diagonal Pad\'e approximant for corresponding power function. Rigorous error analysis is carried out and sharp error bounds are presented which show that the scheme is robust for $\alpha\rightarrow 0^+$ and $\alpha \rightarrow 1^-$. The cost of the proposed algorithm is solving some elliptic problems. Since the approach is exponentially convergent with respect to the number of solves, it is very efficient. Some numerical tests are given to confirm our theoretical analysis and the robustness of the algorithm.

[3]  arXiv:2206.15167 [pdf, ps, other]

Title: Convergence Analysis of Dirichlet Energy Minimization for Spherical Conformal Parameterizations

Authors: Wei-Hung Liao, Tsung-Ming Huang, Wen-Wei Lin, Mei-Heng Yueh

Comments: 29 pages

Subjects: Numerical Analysis (math.NA)

In this paper, we first derive a theoretical basis for spherical conformal parameterizations between a simply connected closed surface $\mathcal{S}$ and a unit sphere $\mathbb{S}^2$ by minimizing the Dirichlet energy on $\overline{\mathbb{C}}$ by stereographic projection. The Dirichlet energy can be rewritten as the sum of the energies associated with the southern and northern hemispheres and can be decreased under an equivalence relation by alternatingly solving the corresponding Laplacian equations. Based on this theoretical foundation, we develop a modified Dirichlet energy minimization with nonequivalence deflation for the computation of the spherical conformal parameterization between $\mathcal{S}$ and $\mathbb{S}^2$. In addition, under some mild conditions, we verify the asymptotically R-linear convergence of the proposed algorithm. Numerical experiments on various benchmarks confirm that the assumptions for convergence always hold and indicate the efficiency, reliability and robustness of the developed modified Dirichlet energy minimization.

Cross-lists for Fri, 1 Jul 22

[4]  arXiv:2206.15051 (cross-list from cs.LG) [pdf, other]

Title: Group-invariant tensor train networks for supervised learning

Authors: Brent Sprangers, Nick Vannieuwenhoven

Subjects: Machine Learning (cs.LG); Numerical Analysis (math.NA)

Invariance has recently proven to be a powerful inductive bias in machine learning models. One such class of predictive or generative models are tensor networks. We introduce a new numerical algorithm to construct a basis of tensors that are invariant under the action of normal matrix representations of an arbitrary discrete group. This method can be up to several orders of magnitude faster than previous approaches. The group-invariant tensors are then combined into a group-invariant tensor train network, which can be used as a supervised machine learning model. We applied this model to a protein binding classification problem, taking into account problem-specific invariances, and obtained prediction accuracy in line with state-of-the-art deep learning approaches.

[5]  arXiv:2206.15397 (cross-list from cs.LG) [pdf, ps, other]

Title: Randomized K-FACs: Speeding up K-FAC with Randomized Numerical Linear Algebra

Authors: Constantin Octavian Puiu

Comments: First version: 12 pages, 2 figures, 1 table

Subjects: Machine Learning (cs.LG); Numerical Analysis (math.NA); Machine Learning (stat.ML)

K-FAC is a successful tractable implementation of Natural Gradient for Deep Learning, which nevertheless suffers from the requirement to compute the inverse of the Kronecker factors (through an eigen-decomposition). This can be very time-consuming (or even prohibitive) when these factors are large. In this paper, we theoretically show that, owing to the exponential-average construction paradigm of the Kronecker factors that is typically used, their eigen-spectrum must decay. We show numerically that in practice this decay is very rapid, leading to the idea that we could save substantial computation by only focusing on the first few eigen-modes when inverting the Kronecker-factors. Randomized Numerical Linear Algebra provides us with the necessary tools to do so. Numerical results show we obtain $\approx2.5\times$ reduction in per-epoch time and $\approx3.3\times$ reduction in time to target accuracy. We compare our proposed K-FAC sped-up versions with a more computationally efficient NG implementation, SENG, and observe we perform on par with it.

Replacements for Fri, 1 Jul 22

[6]  arXiv:2006.15670 (replaced) [pdf, other]

Title: Simplest random walk for approximating Robin boundary value problems and ergodic limits of reflected diffusions

Authors: B. Leimkuhler, A. Sharma, M.V. Tretyakov

Comments: The Annals of Applied Probability (to appear)

Subjects: Numerical Analysis (math.NA); Probability (math.PR)

[7]  arXiv:2103.04903 (replaced) [pdf, other]

Title: A novel relaxation scheme for the numerical approximation of Schrödinger-Poisson type systems

Authors: Agissilaos Athanassoulis, Theodoros Katsaounis, Irene Kyza, Stephen Metcalfe

Comments: 17pages, 10 figures

Subjects: Numerical Analysis (math.NA)

[8]  arXiv:2110.13604 (replaced) [pdf, other]

Title: Finite Element Approximation of Large-Scale Isometric Deformations of Parametrized Surfaces

Authors: Martin Rumpf, Stefan Simon, Christoph Smoch

Subjects: Numerical Analysis (math.NA)

[9]  arXiv:2204.00984 (replaced) [pdf, other]

Title: Stability of the Minimum Energy Path

Authors: Xuanyu Liu, Huajie Chen, Christoph Ortner

Subjects: Numerical Analysis (math.NA)

[10]  arXiv:2205.06060 (replaced) [pdf, other]

Title: $H^1$-stability of an L2-type method on general nonuniform meshes for subdiffusion equation

Authors: Chaoyu Quan, Xu Wu

Subjects: Numerical Analysis (math.NA)

[11]  arXiv:2206.01727 (replaced) [pdf, ps, other]

Title: New Progress in Classic Area: Polynomial Root-squaring and Root-finding

Authors: Victor Y. Pan

Comments: 21 pages

Subjects: Numerical Analysis (math.NA)

[12]  arXiv:2206.06855 (replaced) [pdf, ps, other]

Title: A new convergence proof for approximations of the Stefan problem

Authors: Robert Eymard (LAMA), Thierry Gallouët (I2M)

Subjects: Numerical Analysis (math.NA)

[13]  arXiv:2206.08493 (replaced) [pdf, other]

Title: Nonconforming finite elements for the Brinkman and $-\text{curl}Δ \text{curl}$ problems on cubical meshes

Authors: Qian Zhang, Min Zhang, Zhimin Zhang

Subjects: Numerical Analysis (math.NA)

[14]  arXiv:2206.09425 (replaced) [pdf, other]

Title: Semi-implicit high resolution numerical scheme for conservation laws

Authors: Peter Frolkovič, Michal Žeravý

Comments: Slightly revised version that is submitted to AMC journal

Subjects: Numerical Analysis (math.NA)

[15]  arXiv:2206.12901 (replaced) [pdf, other]

Title: Noise-aware Physics-informed Machine Learning for Robust PDE Discovery

Authors: Pongpisit Thanasutives, Takeshi Morita, Masayuki Numao, Ken-ichi Fukui

Comments: 13 pages, 8 figures. This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice

Subjects: Numerical Analysis (math.NA); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Computational Physics (physics.comp-ph)

[16]  arXiv:2206.13879 (replaced) [pdf, ps, other]

Title: Optimal analysis of finite element methods for the stochastic Stokes equations

Authors: Buyang Li, Shu Ma, Weiwei Sun

Subjects: Numerical Analysis (math.NA)

[17]  arXiv:2002.12873 (replaced) [pdf, ps, other]

Title: Federated Over-Air Subspace Tracking from Incomplete and Corrupted Data

Authors: Praneeth Narayanamurthy, Namrata Vaswani, Aditya Ramamoorthy

Comments: To appear in IEEE Transactions on Signal Processing. changes to writing; more general result provided from which previous result follows as special case

Subjects: Machine Learning (cs.LG); Information Theory (cs.IT); Numerical Analysis (math.NA); Machine Learning (stat.ML)

[18]  arXiv:2110.02457 (replaced) [pdf, other]

Title: GDA-AM: On the effectiveness of solving minimax optimization via Anderson Acceleration

Authors: Huan He, Shifan Zhao, Yuanzhe Xi, Joyce C Ho, Yousef Saad

Comments: 31 Pages, ICLR, minimax, Anderson Acceleration

Subjects: Machine Learning (cs.LG); Numerical Analysis (math.NA)

[19]  arXiv:2206.09411 (replaced) [pdf, other]

Title: A Stirling-type formula for the distribution of the length of longest increasing subsequences, applied to finite size corrections to the random matrix limit

Authors: Folkmar Bornemann

Comments: 25 pages, 6 figures, 3 tables; variance discussed in new {\S}4.4; simplifications in {\S}3.2: derived kernels have finite rank; improved numerical stabilization in {\S}3.3; associated table of exact values has been extended for up to $n=1000$

Subjects: Combinatorics (math.CO); Mathematical Physics (math-ph); Numerical Analysis (math.NA); Probability (math.PR)

[20]  arXiv:2206.13873 (replaced) [pdf, other]

Title: High-order Lohner-type algorithm for rigorous computation of Poincaré maps in systems of Delay Differential Equations with several delays

Authors: Robert Szczelina, Piotr Zgliczyński

Subjects: Dynamical Systems (math.DS); Numerical Analysis (math.NA)

• New submissions
• Cross-lists
• Replacements