Adaptive shrinkage of singular values
From MaRDI portal
Abstract: To recover a low rank structure from a noisy matrix, truncated singular value decomposition has been extensively used and studied. Recent studies suggested that the signal can be better estimated by shrinking the singular values. We pursue this line of research and propose a new estimator offering a continuum of thresholding and shrinking functions. To avoid an unstable and costly cross-validation search, we propose new rules to select two thresholding and shrinking parameters from the data. In particular we propose a generalized Stein unbiased risk estimation criterion that does not require knowledge of the variance of the noise and that is computationally fast. A Monte Carlo simulation reveals that our estimator outperforms the tested methods in terms of mean squared error on both low-rank and general signal matrices across different signal to noise ratio regimes. In addition, it accurately estimates the rank of the signal when it is detectable.
Recommendations
- Generalized SURE for optimal shrinkage of singular values in low-rank matrix denoising
- Adaptive higher-order spectral estimators
- A singular value \(p\)-shrinkage thresholding algorithm for low rank matrix recovery
- Matrix estimation by universal singular value thresholding
- Bootstrap-based regularization for low-rank matrix estimation
Cites work
- scientific article; zbMATH DE number 845714 (Why is no real title available?)
- scientific article; zbMATH DE number 6458321 (Why is no real title available?)
- A Singular Value Thresholding Algorithm for Matrix Completion
- Asymptotics of sample eigenstructure for a large dimensional spiked covariance model
- Bi-cross-validation of the SVD and the nonnegative matrix factorization
- Eigenvalues of large sample covariance matrices of spiked population models
- Equivariant and scale-free Tucker decomposition models
- Estimation of the mean of a multivariate normal distribution
- Global Image Denoising
- Ideal spatial adaptation by wavelet shrinkage
- Matrix estimation by universal singular value thresholding
- Minimax risk of matrix denoising by singular value thresholding
- Model Averaging and Dimension Selection for the Singular Value Decomposition
- Nonlinear shrinkage estimation of large-dimensional covariance matrices
- On the distribution of the largest eigenvalue in principal components analysis
- Optimal Shrinkage of Singular Values
- Partitioning of interaction in analysis of variance
- Practical approaches to principal component analysis in the presence of missing values
- Reconstruction of a low-rank matrix in the presence of Gaussian noise
- Reduced rank regression via adaptive nuclear norm penalization
- Regularised PCA to denoise and visualise data
- Robust Statistics
- Robust principal component analysis?
- Selecting the number of components in principal component analysis using cross-validation approximations
- Sharp Oracle Inequalities for High-Dimensional Matrix Prediction
- Smooth blockwise iterative thresholding: a smooth fixed point estimator based on the likelihood's block gradient
- Smoothing noisy data with spline functions: Estimating the correct degree of smoothing by the method of generalized cross-validation
- Spectral regularization algorithms for learning large incomplete matrices
- The Adaptive Lasso and Its Oracle Properties
- The Optimal Hard Threshold for Singular Values is <inline-formula> <tex-math notation="TeX">\(4/\sqrt {3}\) </tex-math></inline-formula>
- The sparsity and bias of the LASSO selection in high-dimensional linear regression
- Unbiased Risk Estimates for Singular Value Thresholding and Spectral Estimators
Cited in
(17)- Hierarchical nuclear norm penalization for multi-view data integration
- Reconstruction of a low-rank matrix in the presence of Gaussian noise
- Optimal singular value shrinkage for operator norm loss: extending to non-square matrices
- Generalized SURE for optimal shrinkage of singular values in low-rank matrix denoising
- Ridge-type linear shrinkage estimation of the mean matrix of a high-dimensional normal distribution
- Smooth singular value thresholding algorithm for low-rank matrix completion problem
- Sparse Single Index Models for Multivariate Responses
- An adaptive singular value shrinkage for estimation problem of low-rank matrix mean with unknown covariance matrix
- Adaptive singular value shrinkage estimate for low rank tensor denoising
- Imputation of Mixed Data With Multilevel Singular Value Decomposition
- Multiple imputation for continuous variables using a Bayesian principal component analysis
- Accelerating patch-based low-rank image restoration using kd-forest and Lanczos approximation
- Channel estimation for finite scatterers massive multi-user MIMO system
- Adaptive higher-order spectral estimators
- MIMCA: multiple imputation for categorical variables with multiple correspondence analysis
- \textit{ScreeNOT}: exact MSE-optimal singular value thresholding in correlated noise
- Selecting Regularization Parameters for Nuclear Norm--Type Minimization Problems
This page was built for publication: Adaptive shrinkage of singular values
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q294253)
