AI Paper Atlas

Research map / Bandits and online learning / Multi-armed bandits

Finding All $\epsilon$-Good Arms in Stochastic Bandits

Blake Mason, Lalit Jain, Ardhendu Tripathy, Robert Nowak

NeurIPS 2020 · Bandits and online learning › Multi-armed bandits · approach: Best-arm identification in linear bandits

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
use (ST)2 instead of APT oracle (threshold given) for recovering all epsilon-good arms(ST)2 · NYCCC Contest 651, eps=0.1F1 scoreImproved
use (ST)2 instead of UCB, currently used for the contest(ST)2 · NYCCC Contest 651, eps=0.1F1 scoreImproved
misspecify k for LUCB1 by a factor of 2 (k=23 or k=92 instead of 46) (ablation)LUCB1 · NYCCC Contest 651, eps=0.1F1 scoreHurt
use (ST)2 instead of oracle baselines on cancer drug discovery data(ST)2 · ACVRL1 inhibitors, eps=0.8, delta=0.001F1 scoreImproved
use FAREAST instead of (ST)2 when 1/beta^2 >> n/eps^2FAREAST · Simulated Gaussian, delta=0.01, beta=0.01sample complexity vs number of armsImproved
use uniform sampling instead of adaptive (ST)2/FAREASTuniform sampling · Simulated Gaussiansample complexityHurt
use (ST)2 instead of UCB for early precision(ST)2 · NYCCC Contest 651, eps=0.1precisionNo clear effect
use (ST)2 instead of other methods for recall(ST)2 · NYCCC Contest 651, eps=0.1recallNo clear effect

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