Accelerating gradient projection methods for ${\ell }_1$-constrained signal recovery by steplength selection rules

Abstract

We propose a new gradient projection algorithm that compares favorably with the fastest algorithms available to date for ${\ell }_1$-constrained sparse recovery from noisy data, both in the compressed sensing and inverse problem frameworks. The method exploits a line-search along the feasible direction and an adaptive steplength selection based on recent strategies for the alternation of the well-known Barzilai-Borwein rules. The convergence of the proposed approach is discussed and a computational study on both well-conditioned and ill-conditioned problems is carried out for performance evaluations in comparison with five other algorithms proposed in the literature.

arXiv comment11 pages, 4 figures

References in corpus (13)

• Discussion of "Least angle regression" by Efron et al
• Rejoinder to "Least angle regression" by Efron et al
• Least Angle Regression
• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Accelerated Projected Gradient Method for Linear Inverse Problems with Sparsity Constraints
• Tomographic inversion using ${\ell }_1$-norm regularization of wavelet coefficients
• On the performance of algorithms for the minimization of ${\ell }_1$-penalized functionals

Cited by in corpus (11)

• NESTA: A Fast and Accurate First-order Method for Sparse Recovery
• LaBonte's method revisited: An effective steepest descent method for micromagnetic energy minimization
• Proximal Splitting Methods in Signal Processing
• Efficient deconvolution methods for astronomical imaging: algorithms and IDL-GPU codes
• On the convergence of a linesearch based proximal-gradient method for nonconvex optimization
• Nonparametric sparsity and regularization
• A cyclic block coordinate descent method with generalized gradient projections
• Convolution based smooth approximations to the absolute value function with application to non-smooth regularization
• A projected gradient method for $\alpha {\ell }_1-\beta {\ell }_2$ sparsity regularization
• A Barzilai-Borwein $l_1$-Regularized Least Squares Algorithm for Compressed Sensing
• Continuous Time Locally Stationary Wavelet Processes