AI Paper Atlas

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

Spectral bandits

Tomáš Kocák, Rémi Munos, Branislav Kveton, Shipra Agrawal, Michal Valko

JMLR 2020 · Bandits and online learning › Online convex optimization and games · approach: Best-of-both-worlds bandit algorithms

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
make the reward function smoother (fewer nonzero eigenvector components: 500 -> 5) (ablation)SpectralTS · BA graph 500 nodes, T=100cumulative regret: 123.79 -> 7.99Improved
make the reward function smoother (500 -> 5 nonzero components) (ablation)SpectralUCB · BA graph 500 nodes, T=100cumulative regret: 130.54 -> 3.05Improved
use graph-spectral algorithms instead of LinearTS/LinUCBSpectralTS and SpectralUCB · Erdos-Renyi, lattice, Barabasi-Albert graphs (N=500, T=100)cumulative regretImproved
use lazy updates for UCB computation (ablation)SpectralUCB · running time vs number of actionscomputational timeImproved
use only first L=200 (10%) eigenvectors instead of all 2000 (ablation)SpectralUCB · MovieLens single user, N=2019cumulative regret and running timeNo clear effect
use spectral algorithms instead of linear onesSpectralTS and SpectralUCB · MovieLens (400 users, 618 movies), T=500cumulative regretImproved

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