What type of study is this?

This is a Experimental Study study.

September 20, 2025

Problem-dependent Regret for Lexicographic Multi-Armed Bandits with Adversarial Corruptions

View Full Paper

Ask AI

Bookmark

Authors

BXBo XueCity University of Hong Kong XLXi LinCity University of Hong Kong YWYuanyu WanNingbo University

Discussion

Member takes

Implication

Research proposes algorithms improving regret bounds in lexicographic multi-armed bandits, highlighting adversarial corruptions and varying objectives.

Key Points

The proposed algorithm shows a problem-dependent regret bound of O(∑(log T / Δⁱ(a) + C)) for adversarial corruptions.
When C=0, the algorithm achieves near-optimal regret bounds compared to traditional single-objective approaches.
A second algorithm is introduced that does not require knowledge of C, albeit with a less favorable regret bound.
Experiments on synthetic and real-world datasets validate the effectiveness of these algorithms in practical scenarios.

Cite This Study

Xue et al. (2025) studied this question.

synapsesocial.com/papers/68d469d631b076d99fa66fd4 https://doi.org/10.24963/ijcai.2025/754

View Full Paper

Ask AI

Bookmark

Implication

Research proposes algorithms improving regret bounds in lexicographic multi-armed bandits, highlighting adversarial corruptions and varying objectives.

Problem-dependent Regret for Lexicographic Multi-Armed Bandits with Adversarial Corruptions

Discussion

Member takes

Implication

Key Points

Cite This Study

Problem-dependent Regret for Lexicographic Multi-Armed Bandits with Adversarial Corruptions

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider