Generalized SURE for Exponential Families: Applications to Regularization
From MaRDI portal
Abstract: Stein's unbiased risk estimate (SURE) was proposed by Stein for the independent, identically distributed (iid) Gaussian model in order to derive estimates that dominate least-squares (LS). In recent years, the SURE criterion has been employed in a variety of denoising problems for choosing regularization parameters that minimize an estimate of the mean-squared error (MSE). However, its use has been limited to the iid case which precludes many important applications. In this paper we begin by deriving a SURE counterpart for general, not necessarily iid distributions from the exponential family. This enables extending the SURE design technique to a much broader class of problems. Based on this generalization we suggest a new method for choosing regularization parameters in penalized LS estimators. We then demonstrate its superior performance over the conventional generalized cross validation approach and the discrepancy method in the context of image deblurring and deconvolution. The SURE technique can also be used to design estimates without predefining their structure. However, allowing for too many free parameters impairs the performance of the resulting estimates. To address this inherent tradeoff we propose a regularized SURE objective. Based on this design criterion, we derive a wavelet denoising strategy that is similar in sprit to the standard soft-threshold approach but can lead to improved MSE performance.
Cited in
(32)- Risk estimators for choosing regularization parameters in ill-posed problems -- properties and limitations
- Automated parameter selection for total variation minimization in image restoration
- Automated data-driven selection of the hyperparameters for total-variation-based texture segmentation
- Automatic parameter selection based on residual whiteness for convex non-convex variational restoration
- Relaxing the Gaussian assumption in shrinkage and SURE in high dimension
- Degrees of freedom for off-the-grid sparse estimation
- On an unsupervised method for parameter selection for the elastic net
- Convergence of regularization methods with filter functions for a regularization parameter chosen with GSURE and mildly ill-posed inverse problems
- Local behavior of sparse analysis regularization: applications to risk estimation
- Effective new methods for automated parameter selection in regularized inverse problems
- Estimating network degree distributions under sampling: an inverse problem, with applications to monitoring social media networks
- The degrees of freedom of partly smooth regularizers
- Iterative choice of the optimal regularization parameter in TV image restoration
- Parametric PSF estimation based on recursive SURE for sparse deconvolution
- On the convergence of recursive SURE for total variation minimization
- SURE-type functionals as criteria for parametric PSF estimation
- A compressed sensing approach to interpolation of fractional Brownian trajectories for a single particle tracking experiment
- Low complexity regularization of linear inverse problems
- Least squares estimation without priors or supervision
- Recursive SURE for iterative reweighted least square algorithms
- Generalized SURE for optimal shrinkage of singular values in low-rank matrix denoising
- Adaptive singular value shrinkage estimate for low rank tensor denoising
- scientific article; zbMATH DE number 7625169 (Why is no real title available?)
- Sampling Based Estimation of In-Degree Distribution for Directed Complex Networks
- Maximum likelihood estimation of regularization parameters in high-dimensional inverse problems: an empirical Bayesian approach. I: Methodology and experiments
- The projected GSURE for automatic parameter tuning in iterative shrinkage methods
- On the convergence rate of projected gradient descent for a back-projection based objective
- Bilevel Methods for Image Reconstruction
- Efficient Bayesian Computation for Low-Photon Imaging Problems
- Regularisation and central limit theorems for an inverse problem in network sampling applications
- Unbiased test error estimation in the Poisson means problem via coupled bootstrap techniques
- Classifier design given an uncertainty class of feature distributions via regularized maximum likelihood and the incorporation of biological pathway knowledge in steady-state phenotype classification
This page was built for publication: Generalized SURE for Exponential Families: Applications to Regularization
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4571543)
