AI Paper Atlas

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

The Lazy Online Subgradient Algorithm is Universal on Strongly Convex Domains

Daron Anderson, Douglas J. Leith

NeurIPS 2021 · Bandits and online learning › Online convex optimization and games · approach: Constrained and unconstrained OCO algorithms · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
decrease exponent r of the mean cost vector (towards uniform cost direction) under noise type 1Online Lazy Gradient Descent · l_1.5 unit ballaverage pseudo-regretHurt
increase r under noise type 2 (noise on last component)Online Lazy Gradient Descent · l_1.5 unit ballaverage pseudo-regretHurt
increase dimension d (r=0, noise type 1)Online Lazy Gradient Descent · l_p ballaverage pseudo-regretImproved
increase dimension d (r=8, noise type 2)Online Lazy Gradient Descent · l_p ballaverage pseudo-regretHurt
decrease p towards 1 (less curvature)Online Lazy Gradient Descent · l_p ball worst casesaverage pseudo-regretHurt
use Greedy Gradient Descent update instead of LazyOnline Gradient Descent · Euclidean unit ballworst-case regret growthHurt
vary noise type (1,2,3) on Schatten ballOnline Lazy Gradient Descent · Schatten unit ballaverage pseudo-regretNo clear effect
increase d / decrease p on Schatten ballOnline Lazy Gradient Descent · Schatten unit ballaverage pseudo-regretMixed

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