Research map / Bandits and online learning / Online convex optimization and games
The Price of Differential Privacy for Online Learning
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use the proposed private FTRL bound instead of the previous private bound | Online linear optimization over the Sphere · Online linear optimization over the Sphere | regret: \tilde{O}\left(\frac{\sqrt{NT}}{\varepsilon}\right) -> O\left(\sqrt{T} + \frac{N \log^2 T}{\varepsilon} \right) | Helped |
| Use the proposed private FTRL bound instead of the previous private bound | Online linear optimization over the Cube · Online linear optimization over the Cube | regret: \tilde{O}\left(\frac{\sqrt{NT}}{\varepsilon}\right) -> O\left(\sqrt{NT} + \frac{N \log^2 T}{\varepsilon} \right) | Helped |
| Use the proposed private FTRL bound instead of the previous private bound | Online Linear Optimization · Online Linear Optimization | regret: \tilde{O}\left(\frac{\sqrt{T}}{\varepsilon}\right) -> O\left( \sqrt{T} + \frac{\log^2 T}{\varepsilon} \right) | Helped |
| Use the proposed private FTRL bound instead of the previous private bound | Prediction with Expert Advice · Prediction with Expert Advice | regret: \tilde{O}\left(\frac{\sqrt{T\log N}}{\varepsilon}\right) -> O\left(\sqrt{T\log N} + \frac{N\log N \log^2 T}{\varepsilon} | Helped |
| Apply the privacy reduction to a non-private bandit algorithm | Bandit linear optimization · Bandit Linear Optimization | regret: \tilde{O}\left(\frac{T^{\frac{2}{3}}}{\varepsilon}\right) -> \tilde{O}\left(\frac{\sqrt{T}}{\varepsilon}\right) | Helped |
| Use private EXP2 with uniform exploration | Non-stochastic multi-armed bandits · Non-stochastic Multi-armed Bandits | regret: \tilde{O}\left(\frac{NT^{\frac{2}{3}}}{\varepsilon}\right) -> \tilde{O}\left(\frac{\sqrt{TN\log N}}{\varepsilon}\right) | Helped |
Closest papers by meaning
- Connecting Thompson Sampling and UCB: Towards More Efficient Trade-offs Between Privacy and Regret (ICML 2025 · similarity 0.91)
- Improved Differentially Private and Lazy Online Convex Optimization: Lower Regret without Smoothness Requirements (ICML 2024 · similarity 0.90)
- Shuffle and Joint Differential Privacy for Generalized Linear Contextual Bandits (NeurIPS 2026 · similarity 0.90)
- Faster Rates for Private Adversarial Bandits (ICML 2025 · similarity 0.90)
- Private Online Learning via Lazy Algorithms (NeurIPS 2024 · similarity 0.90)
- Revisiting Differentially Private Algorithms for Decentralized Online Learning (ICML 2025 · similarity 0.89)
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.