Searcharxiv⌕ Search

arXiv subjects

Vladimir Druskin

Publications and source records attributed to Vladimir Druskin.

At least 19 recordsLinked to original sources

Adapting the Lanczos algorithm to matrices with almost continuous spectra

We consider approximating $B^T(A+sI)^{-1}B$, where $A\in\mathbb{R}^{n\times n}$ is large, symmetric positive definite, and $B\in\mathbb{R}^{n\times p}$ with $p\ll n$. We focus on the case where $A$ has an almost continuous (dense) spectrum: its eigenvalues fill one or more intervals so densely that Krylov methods cannot resolve them individually. Our target is computing multiple-input multiple-output transfer functions arising from large-scale discretizations of problems with continuous spectral measures, such as linear time-invariant PDEs on unbounded domains. Traditional Krylov methods, such as Lanczos or conjugate gradients, resolve individual eigenvalues of a dense discretization while ignoring the underlying continuous spectral measure these points approximate. We argue it is more efficient to model the operator's inherent branch cut than to exhaustively resolve the artificial point spectrum induced by discretization. We adapt the framework of Kreĭn-Nudelman semi-infinite strings to the block Lanczos algorithm, with parameters chosen adaptively by maximizing the energy absorbed at the string termination relative to the energy stored in the string. This yields a low-rank modification to the block Lanczos matrix, dependent on $\sqrt{s}$, at an additional $O(n)$ cost. We show significant error reductions for large-scale self-adjoint PDE discretizations in unbounded domains, including two- and three-dimensional Maxwell's equations in diffusive regimes. The method is especially advantageous for computing state-space solutions for wave propagation, specifically for 2D wave and 3D Maxwell operators. Replacing the conventional Lanczos spectral decomposition with the continuous Kreĭn-Nudelman spectrum yields a qualitative improvement in finite-difference approximations, transforming standing-wave artifacts into outgoing propagating waves.

math.NA↗

Data generated internal solutions for the plasma wave equation: error bounds and numerical experiments in two dimensions

We consider the computation of internal solutions for a time domain plasma wave equation with an unknown potential $q$ from boundary response data. The internal solutions are computed by transforming known background snapshots using the Cholesky decomposition of the data-driven Gramian, or mass matrix. It was recently shown that in one dimension these data generated internal solutions converge in $L^2$ at order $\sqrtτ$ for well chosen initial waves. Here we study the internal solution reconstruction in two dimensions, where a multiple input/multiple output (MIMO) setup and a block Gramian are needed, with the number of boundary sources increasing as the time sampling is refined. We show that the general error bound carries over to this setting: the distance between the data generated solutions and the best causal approximation from background snapshots is controlled by the best approximation mass matrix mismatch. Numerical refinement studies on a square domain measure the convergence of the data generated solutions alongside the best causal approximation from the background snapshots, with the relative errors appearing to go to zero at rate $\sqrtτ$, and the absolute errors going to zero in $L^1$. We also show that the solution reconstructions remain accurate for high contrast composite media. Finally, since the mass matrix assembled from noisy response data can fail to be positive definite, regularization is needed; comparing a diagonal shift with an eigenvalue floor, we find the floor more robust, with reconstructions that remain more accurate than the unperturbed background field with noise that is up to ten percent of the root mean square data amplitude.

math.NA↗

A Targeted Quadrature Framework for Simulating Large-Scale 3D Anisotropic Electromagnetic Measurements

We develop a new, efficient, and accurate method to simulate frequency-domain borehole electromagnetic (EM) measurements acquired in the presence of three-dimensional (3D) variations of the anisotropic subsurface conductivity. The method is based on solving the quasi-static Maxwell equations with a goal-oriented finite-volume discretization via block-quadrature reduced-order modeling. Discretization is performed with a Lebedev grid that enables accurate and conservative solutions in the presence of any form of anisotropic electrical conductivity. Likewise, the method makes use of a new effective-medium approximation to locally account for non-conformal boundaries and large contrasts in electrical conductivity, especially in the vicinity of EM sources and receivers. The finite-volume discretization yields a large symmetric linear system of equations, which is reduced to a set of smaller structured problems via block Lanczos recursion. The formulation also enables the efficient calculation of the adjoint solution, which is necessary for gradient-based inversion of the measurements to estimate the associated spatial distribution of electrical conductivity, i.e., to solve the inverse problem. Specific applications and verifications of the new numerical simulation algorithm are considered for the case of borehole ultra-deep azimuthal resistivity measurements (UDAR) typically used for subsurface well geosteering and navigation. We verify the efficiency, robustness, and scalability of this approach using synthetic UDAR measurements acquired in a 3D formation inspired by North-Sea geology. The numerical experiments successfully verify the applicability of our modeling approach to real-time UDAR processing frameworks.

physics.geo-ph↗

Lippmann-Schwinger-Lanczos algorithm for inverse scattering problems with unknown reflectivity and loss distributions: One-dimensional Case

We consider one-dimensional inverse scattering in attenuating media where both the reflectivity and loss distributions are unknown. Mathematically, this corresponds to recovering the coefficients of a damped wave operator, or equivalently, a quadratic operator pencil in the frequency domain. The Lippmann-Schwinger equation maps the unknown reflectivity and loss distribution to the measured scattered data. This mapping is nonlinear, as it requires knowledge of the internal wavefield, which itself depends on the reflectivity and loss distribution. The Lippmann-Schwinger-Lanczos method addresses this nonlinearity by approximating the internal solutions through the lifting of states from a reduced-order model constructed directly from the measured data. In this work, we extend the method to dissipative problems, enabling the approximation of internal partial differential equation (PDE) solutions in media with both reflectivity and loss distributions. We present two complementary constructions of such internal solutions: one based on spectral data and another on frequency-domain measurements over a finite interval. This development establishes a direct link between data-driven reduced-order models for inverse problems and port-Hamiltonian dynamical systems, with reduced models obtained either from the associated spectral measure or via rational approximation. Compared to the Born approximation, which replaces the internal field with the background field, our approach yields more accurate internal reconstructions and enables faster and more robust recovery of the contrast as evidenced by our numerical experiments.

math.NA↗

A Krylov projection algorithm for large symmetric matrices with dense spectra

We consider the approximation of $B^T (A+sI)^{-1} B$ for large s.p.d. $A\in\mathbb{R}^{n\times n}$ with dense spectrum and $B\in\mathbb{R}^{n\times p}$, $p\ll n$. We target the computations of Multiple-Input Multiple-Output (MIMO) transfer functions for large-scale discretizations of problems with continuous spectral measures, such as linear time-invariant (LTI) PDEs on unbounded domains. Traditional Krylov methods, such as the Lanczos or CG algorithm, are known to be optimal for the computation of $(A+sI)^{-1}B$ with real positive $s$, resulting in an adaptation to the distinctively discrete and nonuniform spectra. However, the adaptation is damped for matrices with dense spectra. It was demonstrated in [Zimmerling, Druskin, Simoncini, Journal of Scientific Computing 103(1), 5 (2025)] that averaging Gauß and Gauß-Radau quadratures computed using the block-Lanczos method significantly reduces approximation errors for such problems. Here, we introduce an adaptive Kreĭn-Nudelman extension to the (block) Lanczos recursions, allowing further acceleration at negligible $o(n)$ cost. Similar to the Gauß-Radau quadrature, a low-rank modification is applied to the (block) Lanczos matrix. However, unlike the Gauß-Radau quadrature, this modification depends on $\sqrt{s}$ and can be considered in the framework of the Hermite-Padé approximants, which are known to be efficient for problems with branch-cuts, that can be good approximations to dense spectral intervals. Numerical results for large-scale discretizations of heat-diffusion and quasi-magnetostatic Maxwell's operators in unbounded domains confirm the efficiency of the proposed approach.

math.NA↗

Monotonicity, bounds and extrapolation of Block-Gauss and Gauss-Radau quadrature for computing $B^T ϕ(A) B$

In this paper, we explore quadratures for the evaluation of $B^T ϕ(A) B$ where $A$ is a symmetric positive-definite (s.p.d.) matrix in $\mathbb{R}^{n \times n}$, $B$ is a tall matrix in $\mathbb{R}^{n \times p}$, and $ϕ(\cdot)$ represents a matrix function that is regular enough in the neighborhood of $A$'s spectrum, e.g., a Stieltjes or exponential function. These formulations, for example, commonly arise in the computation of multiple-input multiple-output (MIMO) transfer functions for diffusion PDEs. We propose an approximation scheme for $B^T ϕ(A) B$ leveraging the block Lanczos algorithm and its equivalent representation through Stieltjes matrix continued fractions. We extend the notion of Gauss-Radau quadrature to the block case, facilitating the derivation of easily computable error bounds. For problems stemming from the discretization of self-adjoint operators with a continuous spectrum, we obtain sharp estimates grounded in potential theory for Padé approximations and justify averaging algorithms at no added computational cost. The obtained results are illustrated on large-scale examples of 2D diffusion and 3D Maxwell's equations and a graph from the SNAP repository. We also present promising experimental results on convergence acceleration via random enrichment of the initial block $B$.

math.NA↗

Regularized Reduced Order Lippman-Schwinger-Lanczos Method for Inverse Scattering Problems in the Frequency Domain

Inverse scattering has a broad applicability in quantum mechanics, remote sensing, geophysical, and medical imaging. This paper presents a robust direct reduced order model (ROM) method for solving inverse scattering problems based on an efficient approximation of the resolvent operator regularizing the Lippmann-Schwinger-Lanczos (LSL) algorithm. We show that the efficiency of the method relies upon the weak dependence of the orthogonalized basis on the unknown potential in the Schrödinger equation by demonstrating that the Lanczos orthogonalization is equivalent to performing Gram-Schmidt on the ROM time snapshots. We then develop the LSL algorithm in the frequency domain with two levels of regularization. We show that the same procedure can be extended beyond the Schrödinger formulation to the Helmholtz equation, e.g., to imaging the conductivity using diffusive electromagnetic fields in conductive media with localized positive conductivity perturbations. Numerical experiments for Helmholtz and Schrödinger problems show that the proposed bi-level regularization scheme significantly improves the performance of the LSL algorithm, allowing for good reconstructions with noisy data and large data sets.

math.NA↗

Solving inverse scattering problems via reduced-order model embedding procedures

We present a reduced-order model (ROM) methodology for inverse scattering problems in which the reduced-order models are data-driven, i.e. they are constructed directly from data gathered by sensors. Moreover, the entries of the ROM contain localised information about the coefficients of the wave equation. We solve the inverse problem by embedding the ROM in physical space. Such an approach is also followed in the theory of ``optimal grids,'' where the ROMs are interpreted as two-point finite-difference discretisations of an underlying set of equations of a first-order continuous system on this special grid. Here, we extend this line of work to wave equations and introduce a new embedding technique, which we call Krein embedding, since it is inspired by Krein's seminal work on vibrations of a string. In this embedding approach, an adaptive grid and a set of medium parameters can be directly extracted from a ROM and we show that several limitations of optimal grid embeddings can be avoided. Furthermore, we show how Krein embedding is connected to classical optimal grid embedding and that convergence results for optimal grids can be extended to this novel embedding approach. Finally, we also briefly discuss Krein embedding for open domains, that is, semi-infinite domains that extend to infinity in one direction.

math.NA↗

Model order reduction of layered waveguides via rational Krylov fitting

Rational approximation recently emerged as an efficient numerical tool for the solution of exterior wave propagation problems. Currently, this technique is limited to wave media which are invariant along the main propagation direction. We propose a new model order reduction-based approach for compressing unbounded waveguides with layered inclusions. It is based on the solution of a nonlinear rational least squares problem using the RKFIT method. We show that approximants can be converted into an accurate finite difference representation within a rational Krylov framework. Numerical experiments indicate that RKFIT computes more accurate grids than previous analytic approaches and even works in the presence of pronounced scattering resonances. Spectral adaptation effects allow for finite difference grids with dimensions near or even below the Nyquist limit.

math.NA↗

On extension of the data driven ROM inverse scattering framework to partially nonreciprocal arrays

Data-driven reduced order models (ROMs) recently emerged as powerful tool for the solution of inverse scattering problems. The main drawback of this approach is that it was limited to the measurement arrays with reciprocally collocated transmitters and receivers, that is, square symmetric matrix (data) transfer functions. To relax this limitation, we use our previous work [14], where the ROMs were combined with the Lippmann-Schwinger integral equation to produce a direct nonlinear inversion method. In this work we extend this approach to more general transfer functions, including those that are non-symmetric, e.g., obtained by adding only receivers or sources. The ROM is constructed based on the symmetric subset of the data and is used to construct all internal solutions. Remaining receivers are then used directly in the Lippmann-Schwinger equation. We demonstrate the new approach on a number of 1D and 2D examples with non-reciprocal arrays, including a single input/multiple outputs (SIMO) inverse problem, where the data is given by just a single-row matrix transfer function.

math.NA↗

Distance preserving model order reduction of graph-Laplacians and cluster analysis

Graph-Laplacians and their spectral embeddings play an important role in multiple areas of machine learning. This paper is focused on graph-Laplacian dimension reduction for the spectral clustering of data as a primary application. Spectral embedding provides a low-dimensional parametrization of the data manifold which makes the subsequent task (e.g., clustering) much easier. However, despite reducing the dimensionality of data, the overall computational cost may still be prohibitive for large data sets due to two factors. First, computing the partial eigendecomposition of the graph-Laplacian typically requires a large Krylov subspace. Second, after the spectral embedding is complete, one still has to operate with the same number of data points. For example, clustering of the embedded data is typically performed with various relaxations of k-means which computational cost scales poorly with respect to the size of data set. In this work, we switch the focus from the entire data set to a subset of graph vertices (target subset). We develop two novel algorithms for such low-dimensional representation of the original graph that preserves important global distances between the nodes of the target subset. In particular, it allows to ensure that target subset clustering is consistent with the spectral clustering of the full data set if one would perform such. That is achieved by a properly parametrized reduced-order model (ROM) of the graph-Laplacian that approximates accurately the diffusion transfer function of the original graph for inputs and outputs restricted to the target subset. Working with a small target subset reduces greatly the required dimension of Krylov subspace and allows to exploit the conventional algorithms (like approximations of k-means) in the regimes when they are most robust and efficient.

cs.LG↗

A reduced order model approach to inverse scattering in lossy layered media

We introduce a reduced order model (ROM) methodology for inverse electromagnetic wave scattering in layered lossy media, using data gathered by an antenna which generates a probing wave and measures the time resolved reflected wave. We recast the wave propagation problem as a passive infinite-dimensional dynamical system, whose transfer function is expressed in terms of the measurements at the antenna. The ROM is a low-dimensional dynamical system that approximates this transfer function. While there are many possible ROM realizations, we are interested in one that preserves passivity and in addition is: (1) data driven (i.e., is constructed only from the measurements) and (2) it consists of a matrix with special sparse algebraic structure, whose entries contain spatially localized information about the unknown dielectric permittivity and electrical conductivity of the layered medium. Localized means in the intervals of a special finite difference grid. The main result of the paper is to show with analysis and numerical simulations that these unknowns can be extracted efficiently from the ROM.

math.DS↗

Lippmann-Schwinger-Lanczos algorithm for inverse scattering problems

Data-driven reduced order models (ROMs) are combined with the Lippmann-Schwinger integral equation to produce a direct nonlinear inversion method. The ROM is viewed as a Galerkin projection and is sparse due to Lanczos orthogonalization. Embedding into the continuous problem, a data-driven internal solution is produced. This internal solution is then used in the Lippmann-Schwinger equation, thus making further iterative updates unnecessary. We show numerical experiments for spectral domain domain data for which our inversion is far superior to the Born inversion and works as well as when the true internal solution is known.

math.NA↗

Reduced Order Model Approach to Inverse Scattering

We study an inverse scattering problem for a generic hyperbolic system of equations with an unknown coefficient called the reflectivity. The solution of the system models waves (sound, electromagnetic or elastic), and the reflectivity models unknown scatterers embedded in a smooth and known medium. The inverse problem is to determine the reflectivity from the time resolved scattering matrix (the data) measured by an array of sensors. We introduce a novel inversion method, based on a reduced order model (ROM) of an operator called wave propagator, because it maps the wave from one time instant to the next, at interval corresponding to the discrete time sampling of the data. The wave propagator is unknown in the inverse problem, but the ROM can be computed directly from the data. By construction, the ROM inherits key properties of the wave propagator, which facilitate the estimation of the reflectivity. The ROM was introduced previously and was used for two purposes: (1) to map the scattering matrix to that corresponding to the single scattering (Born) approximation and (2) to image i.e., obtain a qualitative estimate of the support of the reflectivity. Here we study further the ROM and show that it corresponds to a Galerkin projection of the wave propagator. The Galerkin framework is useful for proving properties of the ROM that are used in the new inversion method which seeks a quantitative estimate of the reflectivity.

math.NA↗

Reduced order models for spectral domain inversion: Embedding into the continuous problem and generation of internal data

We generate data-driven reduced order models (ROMs) for inversion of the one and two dimensional Schrödinger equation in the spectral domain given boundary data at a few frequencies. The ROM is the Galerkin projection of the Schrödinger operator onto the space spanned by solutions at these sample frequencies. The ROM matrix is in general full, and not good for extracting the potential. However, using an orthogonal change of basis via Lanczos iteration, we can transform the ROM to a block triadiagonal form from which it is easier to extract $q$. In one dimension, the tridiagonal matrix corresponds to a three-point staggered finite-difference system for the Schrödinger operator discretized on a so-called spectrally matched grid which is almost independent of the medium. In higher dimensions, the orthogonalized basis functions play the role of the grid steps. The orthogonalized basis functions are localized and also depend only very weakly on the medium, and thus by embedding into the continuous problem, the reduced order model yields highly accurate internal solutions. That is to say, we can obtain, just from boundary data, very good approximations of the solution of the Schrödinger equation in the whole domain for a spectral interval that includes the sample frequencies. We present inversion experiments based on the internal solutions in one and two dimensions.

math.NA↗

Fast finite-difference convolution for 3D problems in layered media

We developed fast direct solver for 3D Helmholtz and Maxwell equations in layered medium. The algorithm is based on the ideas of cyclic reduction for separable matrices. For the grids with major uniform part (within the survey domain in the problems of geophysical prospecting, for example) and small non-uniform part (PML and coarsening to approximate problems in infinite domain) the computational cost of our approach is $O(N_xN_ylog(N_xN_y)N_z)$. For general non-uniform grids the cost is $O(N^{3/2}_xN^{3/2}_yN_z)$. The first asymptotics coincide with the cost of FFT-based methods, which can be applied for uniform gridding (in x and y) only. Our approach is significantly more efficient compared to the algorithms based on discrete Fourier transform which cost is $O(N^2_xN^2_yN_z)$. The algorithm can be easily extended for solving the elasticity problems as well.

math.NA↗

Robust nonlinear processing of active array data in inverse scattering via truncated reduced order models

We introduce a novel algorithm for nonlinear processing of data gathered by an active array of sensors which probes a medium with pulses and measures the resulting waves. The algorithm is motivated by the application of array imaging. We describe it for a generic hyperbolic system that applies to acoustic, electromagnetic or elastic waves in a scattering medium modeled by an unknown coefficient called the reflectivity. The goal of imaging is to invert the nonlinear mapping from the reflectivity to the array data. Many existing imaging methodologies ignore the nonlinearity i.e., operate under the assumption that the Born (single scattering) approximation is accurate. This leads to image artifacts when multiple scattering is significant. Our algorithm seeks to transform the array data to those corresponding to the Born approximation, so it can be used as a pre-processing step for any linear inversion method. The nonlinear data transformation algorithm is based on a reduced order model defined by a proxy wave propagator operator that has four important properties. First, it is data driven, meaning that it is constructed from the data alone, with no knowledge of the medium. Second, it can be factorized in two operators that have an approximately affine dependence on the unknown reflectivity. This allows the computation of the Fréchet derivative of the reflectivity to the data mapping which gives the Born approximation. Third, the algorithm involves regularization which balances numerical stability and data fitting with accuracy of the order of the standard deviation of additive data noise. Fourth, the algebraic nature of the algorithm makes it applicable to scalar (acoustic) and vectorial (elastic, electromagnetic) wave data without any specific modifications.

math.NA↗

Compressing Large-Scale Wave Propagation Models via Phase-Preconditioned Rational Krylov Subspaces

Rational Krylov subspace (RKS) techniques are well-established and powerful tools for projection-based model reduction of time-invariant dynamic systems. For hyperbolic wavefield problems, such techniques perform well in configurations where only a few modes contribute to the field. RKS methods, however, are fundamentally limited by the Nyquist-Shannon sampling rate, making them unsuitable for the approximation of wavefields in configuration characterized by large travel times and propagation distances, since wavefield responses in such configurations are highly oscillatory in the frequency-domain. To overcome this limitation, we propose to precondition the RKSs by factoring out the rapidly varying frequency-domain field oscillations. The remaining amplitude functions are generally slowly varying functions of source position and spatial coordinate and allow for a significant compression of the approximation subspace. Our one-dimensional analysis together with numerical experiments for large scale 2D acoustic models show superior approximation properties of preconditioned RKS compared with the standard RKS model-order reduction. The preconditioned RKS results in a reduction of the frequency sampling well below the Nyquist-Shannon rate, a weak dependence of the RKS size on the number of inputs and outputs for multiple-input/multiple-output (MIMO) problems, and, most importantly, in a significant coarsening of the finite-difference grid used to generate the RKS. A prototype implementation indicates that the preconditioned RKS algorithm is competitive in the modern high performance computing environment.

math.NA↗