AI Paper Atlas

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

Online Learning with a Hint

Ofer Dekel, arthur flajolet, Nika Haghtalab, Patrick Jaillet

NeurIPS 2017 · Bandits and online learning › Online convex optimization and games · approach: Constrained and unconstrained OCO algorithms · 12 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
Add a weakly correlated directional hint (margin alpha) to online linear optimization over a strongly convex action set, with Algorithm Ahint built on AEXP over virtual lossesonline linear optimization with hints on strongly convex K · theoretical regret analysisworst-case regretImproved
Add a directional hint on a (C,q)-uniformly convex action set with q in (2,3) and a finite hint set (Algorithm Aset)online linear optimization with hints on (C,q)-uniformly convex K · theoretical regret analysisworst-case regretImproved
Provide a directional hint when the action set is a polyhedron (e.g., L1 or L-infinity ball)online linear optimization with hints on polyhedral K · theoretical lower boundworst-case regretNo clear effect
Use a naive strategy that moves the action to the extreme point in direction -v_tonline linear optimization with hints, strongly convex K example · illustrative example (ct=(1,1/2), vt=(0,1), Euclidean ball)regret: linear regret, regret of (√5−1)/2 THurt

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