AI Paper Atlas

Research map / Federated learning and privacy / Differential privacy

Robust and private stochastic linear bandits

Vasileios Charisopoulos, Hossein Esfandiari, Vahab Mirrokni

ICML 2023 · Federated learning and privacy › Differential privacy · approach: Privacy accounting and mechanisms

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 differential privacy (Laplace noise on rewards, LDP model M1) to robust batched arm eliminationrobust arm elimination for stochastic linear bandits · theoretical regret bound (Theorem 3.4)expected regretHurt
Use robust spectral-filtering least-squares estimator instead of plain least squares when rewards are corrupted with probability alphaarm elimination with batched observations · theoretical regret boundexpected regretMixed
Switch privacy model from M1 (per-reward clients) to M2 (per-action client aggregating rewards, Laplace noise on averages)private robust arm elimination · theoretical regret bounddependence on eps_privImproved
Compare to Bogunovic et al. 2021 robust phased elimination regretrobust stochastic linear bandits without privacy · theoretical regret comparisonregret in dimension d: saves a factor of sqrt(d)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