Summary of Fairness Of Exposure in Online Restless Multi-armed Bandits, by Archit Sood et al.
Fairness of Exposure in Online Restless Multi-armed Bandits
by Archit Sood, Shweta Jain, Sujit Gujar
First submitted to arxiv on: 9 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Machine Learning (stat.ML)
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 In this paper, researchers tackle restless multi-armed bandits (RMABs), a type of Markovian decision-making problem where each arm has its own transition dynamics. While solutions exist for both offline and online cases, they don’t consider the distribution of pulls among arms. This leads to unfairness, where some arms aren’t exposed enough. The authors focus on the online scenario, proposing the first fair RMAB framework, where each arm receives pulls proportionate to its merit. They define merit as a function of an arm’s stationary reward distribution and prove that their algorithm achieves sublinear fairness regret in single-pull scenarios. Empirical results also demonstrate good performance in multi-pull scenarios. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about a special type of problem where you need to make decisions based on uncertain information. The goal is to find the best way to do this without being unfair to one option over another. Currently, there are solutions for when all the information is known beforehand or after it’s been decided, but these don’t work well in real-life situations where everything is constantly changing. In this paper, researchers propose a new approach that considers fairness and prove it works well in certain situations. |
