Adaptive restart of accelerated gradient methods under local quadratic growth condition

DOI10.1093/IMANUM/DRZ007MaRDI QIDQ5243267zbMATH OpenOpenAlexWikidataFDO

Authors Olivier Fercoq, Zheng Qu

Publication date 18 November 2019

Published in IMA Journal of Numerical Analysis (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1709.02300

zbMATH Keywords

accelerated gradient descentrestartingquadratic growth conditionunknown error bound

Mathematics Subject Classification ID

Numerical analysis (65-XX)

Recommendations

• Adaptive restart for accelerated gradient schemes
• Adaptive restart of the optimized gradient method for convex optimization
• Accelerated additive Schwarz methods for convex optimization with adaptive restart
• Restart of Accelerated First-Order Methods With Linear Convergence Under a Quadratic Functional Growth Condition
• Accelerate stochastic subgradient method by leveraging local growth condition
• Restarting the accelerated coordinate descent method with a rough strong convexity estimate
• scientific article; zbMATH DE number 4099037
• On restart procedures for the conjugate gradient method
• An adaptive accelerated proximal gradient method and its homotopy continuation for sparse optimization
• Stochastic optimization with adaptive restart: a framework for integrated local and global learning

Cited in

(20)

• Parameter-free FISTA by adaptive restart and backtracking
• On the complexity analysis of the primal solutions for the accelerated randomized dual coordinate ascent
• An inexact proximal augmented Lagrangian framework with arbitrary linearly convergent inner solver for composite convex optimization
• Quadratic error bound of the smoothed gap and the restarted averaged primal-dual hybrid gradient
• Lippmann-Schwinger solvers for the computational homogenization of materials with pores
• Fast gradient methods for uniformly convex and weakly smooth problems
• Practical perspectives on symplectic accelerated optimization
• Universal intermediate gradient method for convex problems with inexact oracle
• A simple nearly optimal restart scheme for speeding up first-order methods
• Faster first-order primal-dual methods for linear programming using restarts and sharpness
• Nearly optimal first-order methods for convex optimization under gradient norm measure: an adaptive regularization approach
• A class of modified accelerated proximal gradient methods for nonsmooth and nonconvex minimization problems
• Convergence rates of an inertial gradient descent algorithm under growth and flatness conditions
• A review of nonlinear FFT-based computational homogenization methods
• Differentially Private Accelerated Optimization Algorithms
• Sharpness, restart, and acceleration
• A piecewise conservative method for unconstrained convex optimization
• Restarting the accelerated coordinate descent method with a rough strong convexity estimate
• FISTA is an automatic geometrically optimized algorithm for strongly convex functions
• Perseus: a simple and optimal high-order method for variational inequalities

This page was built for publication: Adaptive restart of accelerated gradient methods under local quadratic growth condition

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5243267)