REVIEW 2 cited by
Decentralized Bilevel Optimization
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Decentralized Bilevel Optimization
read the original abstract
Bilevel optimization has been successfully applied to many important machine learning problems. Algorithms for solving bilevel optimization have been studied under various settings. In this paper, we study the nonconvex-strongly-convex bilevel optimization under a decentralized setting. We design decentralized algorithms for both deterministic and stochastic bilevel optimization problems. Moreover, we analyze the convergence rates of the proposed algorithms in difference scenarios including the case where data heterogeneity is observed across agents. Numerical experiments on both synthetic and real data demonstrate that the proposed methods are efficient.
Forward citations
Cited by 2 Pith papers
-
Optimization under Persistent State-Dependent Bias: Gradient-based Method and Complexity Analysis
Residual Learning, a proposed bilevel gradient method, claims exact convergence under state-dependent analog-hardware bias with rate O~(kappa1*kappa2^4*sigma^2/(mu*K)).
-
Nonconvex Decentralized Stochastic Bilevel Optimization under Heavy-Tailed Noise
The paper introduces D-NSVRGDA, a decentralized normalized variance-reduced method for nonconvex bilevel optimization, and proves the first convergence rate under heavy-tailed noise without gradient clipping.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.