atlas

Research map / Bandits and online learning / Online convex optimization and games

An Ellipsoid Algorithm for Online Convex Optimization

NeurIPS 2025 · Bandits and online learning › Online convex optimization and games · approach: Constrained and unconstrained OCO algorithms · 0 citations (OpenAlex)

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes an ellipsoid-based algorithm for online convex optimization that uses only a separation oracle, avoiding Euclidean projections onto the feasible set.

Problem

Online convex optimization over a convex set accessed via a separation oracle, where projection-based methods like projected OGD require costly Euclidean projections onto the feasible set whenever iterates leave it.

Approach

Replaces Euclidean projections with ellipsoid-method updates driven by separation oracle queries, so the iterates are maintained without projecting onto K, while keeping the optimal O(sqrt(T)) regret order (per the abstract's framing of OGD as the benchmark).

Key results

Experimental setup

Metrics: regret

Baselines: projected online gradient descent

Builds on: online gradient descent ellipsoid method

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy