Abstract:The numerical solution of the generalized eigenvalue problem for a singular matrix pencil is challenging due to the discontinuity of its eigenvalues. Classically, such problems are addressed by first extracting the regular part through the staircase form and then applying a standard solver, such as the QZ algorithm, to that regular part. Recently, several novel approaches have been proposed to transform the singular pencil into a regular pencil by relatively simple randomized modifications. In this work, we analyze three such methods by Hochstenbach, Mehl, and Plestenjak that modify, project, or augment the pencil using random matrices. All three methods rely on the normal rank and do not alter the finite eigenvalues of the original pencil. We show that the eigenvalue condition numbers of the transformed pencils are unlikely to be much larger than the $\delta$-weak eigenvalue condition numbers, introduced by Lotz and Noferini, of the original pencil. This not only indicates favorable numerical stability but also reconfirms that these condition numbers are a reliable criterion for detecting simple finite eigenvalues. We also provide evidence that, from a numerical stability perspective, the use of complex instead of real random matrices is preferable even for real singular matrix pencils and real eigenvalues. As a side result, we provide sharp left tail bounds for a product of two independent random variables distributed with the generalized beta distribution of the first kind or Kumaraswamy distribution.

Random path method with pivoting for computing permanents of matrices

A Hybrid Algorithm for Computing Permanents of Sparse Matrices

Distribution of the number of pivots needed using Gaussian elimination with partial pivoting on random matrices

Randomized Complete Pivoting for Solving Symmetric Indefinite Linear Systems

Approximating Permanent of Random Matrices with Vanishing Mean: Made Better and Simpler

An Algorithmic Proof of Brégman–Minc Theorem

An upper bound for the permanent of (0,1)-matrices

Additive estimates of the permanent using Gaussian fields

On the Efficiency of Random Permutation for ADMM and Coordinate Descent

An analysis on the efficiency of Euler's method for computing the matrix pth root: Analysis on the efficiency of Euler's method for computing the matrix pth root

Analysis of eigenvalue condition numbers for a class of randomized numerical methods for singular matrix pencils

CholeskyQR with Randomization and Pivoting for Tall Matrices (CQRRPT)

Randomized methods for computing joint eigenvalues, with applications to multiparameter eigenvalue problems and root finding

Performance Analysis of Random Competition Strategy in Monte Carlo Algorithms

A Multilevel Monte Carlo Estimator for Matrix Multiplication

Randomized Matrix Approximation to Enhance Regularized Projection Schemes in Inverse Problems

Projected Multilevel Monte Carlo Method for PDE with random input data

CPQR-based randomized algorithms for generalized CUR decompositions

Randomized algorithms for distributed computation of principal component analysis and singular value decomposition

The Road to Deterministic Matrices with the Restricted Isometry Property