Research map / Bandits and online learning / Multi-armed bandits
Stochastic Graphical Bandits with Adversarial Corruptions
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 Elise instead of BARBAR | MAB with adversarially corrupted stochastic rewards · MAB experiments, K = 5 and K = 10 | cumulative regret | Helped |
| Use Elise instead of UCB | MAB with adversarially corrupted stochastic rewards · MAB experiments, K = 5 and K = 10 | cumulative regret | Helped |
| Use Elise instead of Exp3-Set | graphical bandits with adversarially corrupted stochastic rewards · Graphical bandits experiments, p = 0.2 and p = 0.5 | cumulative regret | Helped |
| Use Elise instead of AAE-AlphaSample | graphical bandits with adversarially corrupted stochastic rewards · Graphical bandits experiments, p = 0.2 and p = 0.5 | cumulative regret | Helped |
| Replace BARBAR's expected observation count with the actual observation count when estimating mean rewards (ablation) | Elise · stochastic graphical bandits with adversarial corruptions | regret-bound validity | Helped |
| Use Bernstein inequality instead of the multiplicative Chernoff-Hoeffding inequality in the regret analysis (ablation) | Elise · stochastic graphical bandits with adversarial corruptions | leading constant in the regret bound: 1024 -> 273 | Helped |
| Use a smaller epoch-length constant (ablation) | Elise · MAB experiments | cumulative regret | Helped |
| Restrict exploration to a carefully constructed independent set instead of the full arm set | Elise · stochastic graphical bandits with adversarial corruptions | regret: O(K ln K ln T + KC) -> O(α ln K ln T + αC) | Helped |
Closest papers by meaning
- Stochastic Bandits Robust to Adversarial Attacks (ICLR 2025 · similarity 0.92)
- Adversarial Combinatorial Semi-bandits with Graph Feedback (ICML 2025 · similarity 0.91)
- A Near-optimal, Scalable and Parallelizable Framework for Stochastic Bandits Robust to Adversarial Corruptions and Beyond (NeurIPS 2025 · similarity 0.91)
- Practical Adversarial Attacks on Stochastic Bandits via Fake Data Injection (NeurIPS 2026 · similarity 0.90)
- Sparsity-Agnostic Linear Bandits with Adaptive Adversaries (NeurIPS 2024 · similarity 0.89)
- Bandits with Ranking Feedback (NeurIPS 2024 · 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.