• Corpus ID: 1225017

Efficient evaluation of scaled proximal operators

  title={Efficient evaluation of scaled proximal operators},
  author={Michael P. Friedlander and Gabriel Goh},
Quadratic-support functions [Aravkin, Burke, and Pillonetto; J. Mach. Learn. Res. 14(1), 2013] constitute a parametric family of convex functions that includes a range of useful regularization terms found in applications of convex optimization. We show how an interior method can be used to efficiently compute the proximal operator of a quadratic-support function under different metrics. When the metric and the function have the right structure, the proximal map can be computed with cost nearly… 

Figures and Tables from this paper

Cardinality, Simplex and Proximal Operator ∗

In this report, we consider the proximal operator for the (cid:96) 0 “norm”, which is widely adopted as a sparsity-inducing penalty in e.g. compressed sensing, sparse signal representation and

Adaptive FISTA for Non-convex Optimization

It turns out that in some situations, the proposed algorithm is equivalent to a class of SR1 (identity minus rank 1) proximal quasi-Newton methods, and convergence is proved in a general non-convex setting, and hence, as a byproduct, new convergence guarantees for proximal semi-newton methods are obtained.

On Quasi-Newton Forward-Backward Splitting: Proximal Calculus and Convergence

A framework for quasi-Newton forward--backward splitting algorithms (proximal quasi- newton methods) with a metric induced by diagonal $\pm$ rank-$r$ symmetric positive definite matrices is introduced, which allows for a highly efficient evaluation of the proximal mapping.

The Proximal Bootstrap for Finite-Dimensional Regularized Estimators

  • Jessie Li
  • Mathematics, Computer Science
    AEA Papers and Proceedings
  • 2021
We propose a proximal bootstrap that can consistently estimate the limiting distribution of sqrt(n)-consistent estimators with nonstandardasymptotic distributions in a computationally efficient

Optimization of Graph Total Variation via Active-Set-based Combinatorial Reconditioning

This work proposes a novel adaptive preconditioner driven by a sharp analysis of the local linear convergence rate depending on the "active set" at the current iterate, and shows that nested-forest decomposition of the inactive edges yields a guaranteed locallinear convergence rate.

A New Homotopy Proximal Variable-Metric Framework for Composite Convex Minimization

A new primal-dual-primal framework for implementing proximal Newton methods that has attractive computational features for a subclass of nonsmooth composite convex minimization problems is suggested.

Further properties of the forward–backward envelope with applications to difference-of-convex programming

The forward–backward envelope first introduced in Patrinos and Bemporad is studied and it is demonstrated how to minimize some difference-of-convex regularized least squares problems by minimizing a suitably constructed forward– backward envelope.

ProxSkip: Yes! Local Gradient Steps Provably Lead to Communication Acceleration! Finally!

ProxSkip is a surprisingly simple and provably provably effective method for minimizing the sum of a smooth and an expensive nonsmooth proximable function and offers an effec-tive acceleration of communication complexity.

Generalized self-concordant functions: a recipe for Newton-type methods

The proposed theory provides a mathematical tool to analyze both local and global convergence of Newton-type methods without imposing unverifiable assumptions as long as the underlying functionals fall into the class of generalized self-concordant functions.

One-Step Estimation with Scaled Proximal Methods

We study statistical estimators computed using iterative optimization methods that are not run until completion. Classical results on maximum likelihood estimators (MLEs) assert that a one-step



A quasi-Newton proximal splitting method

Efficient implementations of the proximity calculation for a useful class of functions are described and an elegant quasi-Newton method is applied to acceleration of convex minimization problems, and compares favorably against state-of-the-art alternatives.

Optimizing Costly Functions with Simple Constraints: A Limited-Memory Projected Quasi-Newton Algorithm

An optimization algorithm for minimizing a smooth function over a convex set by minimizing a diagonal plus lowrank quadratic approximation to the function, which substantially improves on state-of-the-art methods for problems such as learning the structure of Gaussian graphical models and Markov random elds.

Proximal Quasi-Newton for Computationally Intensive L1-regularized M-estimators

It is shown that the proximal quasi-Newton method is provably super-linearly convergent, even in the absence of strong convexity, by leveraging a restricted variant of strong Convexity.

IMRO: A Proximal Quasi-Newton Method for Solving ℓ1-Regularized Least Squares Problems

This work presents a proximal quasi-Newton method in which the approximation of the Hessian has the special format of “identity minus rank one” (IMRO) in each iteration, and provides a complexity analysis for variants of IMRO, showing that it matches known best bounds.

Modular Proximal Optimization for Multidimensional Total-Variation Regularization

1D-TV solvers provide the backbone for building more complex (two or higher-dimensional) TV solvers within a modular proximal optimization approach and are provided in an easy to use multi-threaded C++, Matlab and Python library.

Proximal Newton-Type Methods for Minimizing Composite Functions

Newton-type methods for minimizing smooth functions are generalized to handle a sum of two convex functions: a smooth function and a nonsmooth function with a simple proximal mapping, which yields new convergence results for some of these methods.

An Interior-Point Method for Large-Scale $\ell_1$-Regularized Least Squares

A specialized interior-point method for solving large-scale -regularized LSPs that uses the preconditioned conjugate gradients algorithm to compute the search direction and can solve large sparse problems, with a million variables and observations, in a few tens of minutes on a PC.

Fast Newton-type Methods for Total Variation Regularization

This work studies anisotropic (l1-based) TV and also a related l2-norm variant and develops Newton-type methods that outperform the state-of-the-art algorithms for solving the harder task of computing 2- (and higher)-dimensional TV proximity.

An inexact successive quadratic approximation method for L-1 regularized optimization

The inexactness conditions are based on a semi-smooth function that represents a (continuous) measure of the optimality conditions of the problem, and that embodies the soft-thresholding iteration.

Epi-convergent Smoothing with Applications to Convex Composite Functions

Epi-convergence techniques are used to define a notion of epi-smoothing that allows us to tap into the rich variational structure of the subdifferential calculus for nonsmooth, nonconvex, and nonfinite-valued functions.