Research map / Bandits and online learning / Multi-armed bandits
Combinatorial Bandits for Maximum Value Reward Function under Value-Index Feedback
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.
Studies combinatorial bandits where the reward of selecting k arms is the maximum sampled value, under a weak value-index feedback that reveals only the max arm's value and index.
Problem
Combinatorial multi-armed bandits in which an action selects k base arms and the reward is the maximum of their sampled values, observed only through value-index feedback (value and index of the maximum arm) rather than semi-bandit feedback.
Approach
Introduces a value-index feedback structure, weaker than semi-bandit feedback yet slightly stronger than full-bandit feedback, and designs bandit algorithms for maximum-value reward functions under it.
Experimental setup
Metrics: regret
Builds on: combinatorial multi-armed bandits semi-bandit feedback
Closest papers by meaning
- Adversarial Combinatorial Semi-bandits with Graph Feedback (ICML 2025 · similarity 0.92)
- Multi-Armed Bandits With Best-Action Queries (NeurIPS 2026 · similarity 0.91)
- Bandits with Ranking Feedback (NeurIPS 2024 · similarity 0.91)
- On Interpolating Experts and Multi-Armed Bandits (ICML 2024 · similarity 0.91)
- Oracle-Efficient Combinatorial Semi-Bandits (NeurIPS 2025 · similarity 0.91)
- Optimal and Efficient Contextual Combinatorial Semi-bandits with General Function Approximation (NeurIPS 2026 · similarity 0.90)
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.