Singular vectors under random perturbation
From MaRDI portal
Abstract: Computing the first few singular vectors of a large matrix is a problem that frequently comes up in statistics and numerical analysis. Given the presence of noise, exact calculation is hard to achieve, and the following problem is of importance: vskip2mm centerline {it How much a small perturbation to the matrix changes the singular vectors ?} vskip2mm Answering this question, classical theorems, such as those of Davis-Kahan and Wedin, give tight estimates for the worst-case scenario. In this paper, we show that if the perturbation (noise) is random and our matrix has low rank, then better estimates can be obtained. Our method relies on high dimensional geometry and is different from those used an earlier papers.
Recommendations
- Random perturbation of low rank matrices: improving classical bounds
- The singular values and vectors of low rank perturbations of large rectangular random matrices
- Singular vector perturbation under Gaussian noise
- Perturbation of the SVD in the presence of small singular values
- Optimality of singular vector perturbation under maximum norm
Cites work
- Asymptotics of sample eigenstructure for a large dimensional spiked covariance model
- Concentration of norms and eigenvalues of random matrices
- scientific article; zbMATH DE number 2222775 (Why is no real title available?)
- On the concentration of eigenvalues of random symmetric matrices
- Random matrices: The distribution of the smallest singular values
- Spectral algorithms
- The eigenvalues and eigenvectors of finite, low rank perturbations of large random matrices
- The Rotation of Eigenvectors by a Perturbation. III
Cited in
(36)- Random perturbation of low rank matrices: improving classical bounds
- Improved random perturbation intervals of symmetric eigenvalue problem
- Rate-optimal perturbation bounds for singular subspaces with applications to high-dimensional statistics
- Second order accurate distributed eigenvector computation for extremely large matrices
- On the \(\ell^\infty\)-norms of the singular vectors of arbitrary powers of a difference matrix with applications to sigma-delta quantization
- Relative perturbation bounds with applications to empirical covariance operators
- Entrywise eigenvector analysis of random matrices with low expected rank
- Recovery of simultaneous low rank and two-way sparse coefficient matrices, a nonconvex approach
- Random projections as regularizers: learning a linear discriminant from fewer observations than dimensions
- Asymmetry helps: eigenvalue and eigenvector analyses of asymmetrically perturbed low-rank matrices
- 2D and 3D image reconstruction from slice data based on a constrained bilateral smoothing and dynamic mode decomposition
- Understanding singular vectors
- Perturbation of linear forms of singular vectors under Gaussian noise
- A probability argument in favor of ignoring small singular values
- The singular values and vectors of low rank perturbations of large rectangular random matrices
- An \(\ell_{\infty}\) eigenvector perturbation bound and its application
- Unperturbed: spectral analysis beyond Davis-Kahan
- A new random perturbation interval of symmetric eigenvalue problem
- scientific article; zbMATH DE number 7626745 (Why is no real title available?)
- Optimality of singular vector perturbation under maximum norm
- A sharp blockwise tensor perturbation bound for orthogonal iteration
- Perturbation bounds for eigenspaces under a relative gap condition
- Spectral Methods for Passive Imaging: Nonasymptotic Performance and Robustness
- Singular vector perturbation under Gaussian noise
- Signal-plus-noise matrix models: eigenvector deviations and fluctuations
- Perturbation upper bounds for singular subspaces with a kind of heteroskedastic noise and its application in clustering
- Spiked singular values and vectors under extreme aspect ratios
- Euclidean Representation of Low-Rank Matrices and Its Geometric Properties
- Spectral clustering and the high-dimensional stochastic blockmodel
- A stochastic perturbation analysis of the QR decomposition and its applications
- Davis-Kahan theorem under a moderate gap condition
- Curse of dimensionality on persistence diagrams
- Uniform error bound for PCA matrix denoising
- Multivariate Analysis for Multiple Network Data via Semi-Symmetric Tensor PCA
- Community detection in sparse networks via Grothendieck's inequality
- Perturbation of the SVD in the presence of small singular values
This page was built for publication: Singular vectors under random perturbation
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5388973)
