A quasi-Newton proximal splitting method

Abstract

A new result in convex analysis on the calculation of proximity operators in certain scaled norms is derived. We describe efficient implementations of the proximity calculation for a useful class of functions; the implementations exploit the piece-wise linear nature of the dual problem. The second part of the paper applies the previous result to acceleration of convex minimization problems, and leads to an elegant quasi-Newton method. The optimization method compares favorably against state-of-the-art alternatives. The algorithm has extensive applications including signal processing, sparse recovery and machine learning and classification.

References in corpus (1)

• Adaptive Restart for Accelerated Gradient Schemes

Cited by in corpus (29)

• An inertial forward-backward algorithm for monotone inclusions
• Forward-backward quasi-Newton methods for nonsmooth optimization problems
• Composite Self-Concordant Minimization
• Joint reconstruction strategy for structured illumination microscopy with unknown illuminations
• Efficient Solvers for Sparse Subspace Clustering
• The Practicality of Stochastic Optimization in Imaging Inverse Problems
• A General Destriping Framework for Remote Sensing Images Using Flatness Constraint
• Forward-backward truncated Newton methods for convex composite optimization
• SPOQ ${\ell }_p$-Over-${\ell }_q$ Regularization for Sparse Signal Recovery applied to Mass Spectrometry
• On the acceleration of forward-backward splitting via an inexact Newton method
• Sharp MSE Bounds for Proximal Denoising
• Proximal Gradient Algorithms: Applications in Signal Processing
• Statistical Inference for the Population Landscape via Moment Adjusted Stochastic Gradients
• Preconditioned P-ULA for Joint Deconvolution-Segmentation of Ultrasound Images -- Extended Version
• Practical Inexact Proximal Quasi-Newton Method with Global Complexity Analysis
• Efficient evaluation of scaled proximal operators
• PNKH-B: A Projected Newton-Krylov Method for Large-Scale Bound-Constrained Optimization
• Proximal Adam: Robust Adaptive Update Scheme for Constrained Optimization
• A FISTA-type accelerated gradient algorithm for solving smooth nonconvex composite optimization problems
• An Inexact Variable Metric Proximal Point Algorithm for Generic Quasi-Newton Acceleration
• An Inexact Semi-smooth Newton Method on Riemannian Manifolds with Application to Duality-based Total Variation Denoising
• A Proximal Stochastic Quasi-Newton Algorithm
• Generalized Self-Concordant Functions: A Recipe for Newton-Type Methods
• Performance of First- and Second-Order Methods for L1-Regularized Least Squares Problems
• Combinatorial Preconditioners for Proximal Algorithms on Graphs
• Minimizing Oracle-Structured Composite Functions
• Newton acceleration on manifolds identified by proximal-gradient methods
• Trust-Region Methods for Sparse Relaxation
• A Scalable and Extensible Framework for Superposition-Structured Models