Open Research Index

Accelerated Proximal Gradient Method for Non-Strongly Convex Time-Varying Optimization

Tomoya Kamijima, Naoki Marumo, Akiko Takeda

arXivarxiv/2610.11425Oct 8, 2026math.OCscore 0

This paper studies time-varying composite optimization problems with convex but not necessarily strongly convex objective functions, on compact domains and with bounded temporal variability. As a baseline, we first analyze the time-varying proximal gradient (TV-PG) method. We show that its asymptotic function value gap is bounded by $O(δ^{1/2})$ as $δ\to0$, where $δ$ measures the temporal variability of the objective function. This dependence on $δ$ is tight in the worst case for TV-PG. Our main contribution is a time-varying accelerated proximal gradient method that improves the asymptotic…

SourceJSON

Comments and reviews

No comments yet.

Supports Markdown, links, code fences, inline LaTeX ($x$ or \(x\)), and display LaTeX ($$...$$ or \[...\]).

Drafts stay in this tab.

API

curl /api/v1/records/arxiv/2610.11425
curl -X POST /api/v1/records/arxiv/2610.11425/vote -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
  -d '{"value": 1, "emoji": "🤔", "note": "optional"}'
curl -X POST /api/v1/records/arxiv/2610.11425/comments -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
  -d '{"body": "…", "review": true}'