Back to Search Start Over

Bregman Finito/MISO for nonconvex regularized finite sum minimization without Lipschitz gradient continuity

Authors :
Latafat, Puya
Themelis, Andreas
Ahookhosh, Masoud
Patrinos, Panagiotis
Source :
SIAM J Optim 32(3):2230-2262 (2022)
Publication Year :
2021

Abstract

We introduce two algorithms for nonconvex regularized finite sum minimization, where typical Lipschitz differentiability assumptions are relaxed to the notion of relative smoothness. The first one is a Bregman extension of Finito/MISO, studied for fully nonconvex problems when the sampling is randomized, or under convexity of the nonsmooth term when it is essentially cyclic. The second algorithm is a low-memory variant, in the spirit of SVRG and SARAH, that also allows for fully nonconvex formulations. Our analysis is made remarkably simple by employing a Bregman Moreau envelope as Lyapunov function. In the randomized case, linear convergence is established when the cost function is strongly convex, yet with no convexity requirements on the individual functions in the sum. For the essentially cyclic and low-memory variants, global and linear convergence results are established when the cost function satisfies the Kurdyka-\L ojasiewicz property.

Details

Database :
arXiv
Journal :
SIAM J Optim 32(3):2230-2262 (2022)
Publication Type :
Report
Accession number :
edsarx.2102.10312
Document Type :
Working Paper
Full Text :
https://doi.org/10.1137/21M140376X