Summary of Fairness with Exponential Weights, by Stephen Pasteris et al.
Fairness with Exponential Weights
by Stephen Pasteris, Chris Hicks, Vasilios Mavroudis
First submitted to arxiv on: 6 Nov 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 develops a meta-algorithm that converts efficient implementations of Hedge (or discrete Bayesian inference) into algorithms for contextual bandit problems, ensuring exact statistical parity on every trial. This approach guarantees the same asymptotic regret bound as running Exp4 independently for each protected characteristic. The algorithm can handle non-stationarity and estimate the true population from data, making it useful when the true distribution is unknown. The paper’s results are novel and important in their own right. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper creates a way to remove unfairness in some applications by turning algorithms into ones that guarantee fairness on every trial. This approach has the same benefits as running another algorithm separately for each group, but can handle changing conditions and learn from data. The results are new and significant. |
Keywords
» Artificial intelligence » Bayesian inference
