AI Paper Atlas

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

On the Optimality of Perturbations in Stochastic and Adversarial Multi-armed Bandit Problems

Baekjin Kim, Ambuj Tewari

NeurIPS 2019 · Bandits and online learning › Multi-armed bandits · approach: Optimal stochastic bandit algorithms · 3 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
Use Gaussian perturbations with parameter 2 and a matching lower-tail conditionFTPL for stochastic bandits · Stochastic multi-armed banditsinstance-dependent regret: C'' \sum_{\Delta_i>0}\big( {\log(T \Delta_i^2)}/{\Delta_i} + \Delta_i \big)Improved
Use bounded-support perturbations with the prescribed confidence-interval scalingFTPL for stochastic bandits · Stochastic multi-armed banditsinstance-dependent regret: C''' \sum_{\Delta_i>0}\Big( {\log(T)}/{\Delta_i} + \Delta_i \Big)Improved
Replace deterministic UCB choices with Uniform or Rademacher perturbations and a wider confidence intervalUCB1 · Stochastic multi-armed banditsregret: same regret bound as that of UCB1No clear effect
Use bounded perturbations without the extra logarithmic scaling in TFTPL for stochastic bandits · Two-armed bandit exampleregret: linear regretHurt
Tune Gaussian and Double-exponential FTPL perturbationsFTPL · Simulated stochastic bandits with four reward distributionsaverage regretImproved
Use UCB1 rather than RCBUCB1 and RCB · Simulated stochastic bandits with Rademacher rewardsaverage regretHurt
Use RCB with Uniform or Rademacher perturbations rather than UCB1Randomized Confidence Bound · Simulated stochastic bandits with four reward distributionsaverage regretMixed
Seek an exact additive-perturbation equivalent of Tsallis entropy for K >= 4FTPL for adversarial bandits · Adversarial K-armed banditsexistence of equivalent choice probabilities: K \ge 4Hurt
Use Fréchet-type rather than Gumbel-type perturbationsGBPA for adversarial bandits · Adversarial multi-armed banditsexpected block maximum: K^{1/\alpha}Improved

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