Summary of Improved Regret For Bandit Convex Optimization with Delayed Feedback, by Yuanyu Wan and Chang Yao and Mingli Song and Lijun Zhang
Improved Regret for Bandit Convex Optimization with Delayed Feedback
by Yuanyu Wan, Chang Yao, Mingli Song, Lijun Zhang
First submitted to arxiv on: 14 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper investigates bandit convex optimization (BCO) with delayed feedback, where only the loss value of the action is revealed under an arbitrary delay. The authors aim to improve upon previous studies by closing the gap between the existing upper bound and lower bound. They develop a novel algorithm that achieves a regret bound of O(sqrt(n)T^(3/4)+sqrt(dT)) in general. This improvement is significant when d=O((nd)(2/3)T^(1/3)). The authors also demonstrate that their algorithm can achieve tighter bounds for strongly convex functions, with a regret bound of O((nT)(2/3)log^(1/3)(T)+dlog(T)) and O(nsqrt(Tlog(T))+dlog(T)) for unconstrained action sets. This paper contributes to the development of algorithms for BCO problems with delayed feedback. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The researchers study how to make decisions when you can only get feedback on what happened later. They improve upon previous methods by closing a gap between how well different approaches do. Their new algorithm does better than before, especially in certain situations. The paper also shows that their method works well for specific types of problems. |
Keywords
* Artificial intelligence * Optimization
