atlas

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

Logarithmic Regret for Online Control

NeurIPS 2019 · Bandits and online learning › Online convex optimization and games · approach: Dynamic and adaptive regret · 47 citations (OpenAlex)

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's title only (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
Apply OGD to strongly convex surrogate costs over disturbance-action policiesOnline control of linear dynamical systems · Linear dynamical systems with strongly convex costsregret: O(\sqrt{T}) -> \tilde{O}\left(\frac{G^2 W^4}{\alpha \sigma^2} \log^7(T)\right)Helped
Use natural-gradient preconditioning instead of ordinary gradient updatesOnline control with disturbance-action policies · Linear dynamical systems with strongly convex costsregret: \tilde{O}\left(\frac{G W^2}{\alpha \mu} \log^7(T)\right)Mixed
Use natural-gradient updates for spherically symmetric quadratic costsOnline control with disturbance-action policies · Spherically symmetric quadratic costsregret: \tilde{O}\left(\frac{\beta^2 W^2}{\alpha} \log^7(T)\right)Helped

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