Research map / Bandits and online learning / Multi-armed bandits
Pareto Optimal Risk-Agnostic Distributional Bandits with Heavy-Tail Rewards
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes DistLCB, a risk-agnostic distributional bandit algorithm for heavy-tailed rewards that uses 1-Wasserstein deviation inequalities to bound Lipschitz risk measures and achieves asymptotic optimality.
Problem
Multi-armed bandits where the goal is to identify Pareto-optimal arms across multiple risk measures simultaneously, in heavy-tailed reward settings.
Approach
Builds confidence intervals for Lipschitz risk measures from novel deviation inequalities for the 1-Wasserstein distance, enabling a single risk-agnostic algorithm (DistLCB) that handles many risk measures at once, with first lower bounds for risk-measure bandits.
Closest papers by meaning
- Optimal Algorithms for Lipschitz Bandits with Heavy-tailed Rewards (ICML 2019 · similarity 0.91)
- Improved Regret Bounds for Linear Bandits with Heavy-Tailed Rewards (NeurIPS 2025 · similarity 0.90)
- Nash Social Welfare for Multi Armed Bandits: Trajectory-wise Expected and High Probability Regret (NeurIPS 2026 · similarity 0.89)
- Bayesian Regret Minimization in Offline Bandits (ICML 2024 · similarity 0.88)
- Catoni Contextual Bandits are Robust to Heavy-tailed Rewards (ICML 2025 · similarity 0.87)
- p-Mean Regret for Stochastic Bandits (AAAI 2025 · similarity 0.87)
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.