Summary of Intrinsic Fairness-accuracy Tradeoffs Under Equalized Odds, by Meiyu Zhong et al.
Intrinsic Fairness-Accuracy Tradeoffs under Equalized Odds
by Meiyu Zhong, Ravi Tandon
First submitted to arxiv on: 12 May 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Information Theory (cs.IT)
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 This paper explores the tension between fairness and accuracy in machine learning (ML) systems, which are increasingly used in areas like law enforcement, finance, and hiring. The authors present a new theoretical upper bound on the accuracy of ML models that guarantees equalized odds under certain conditions. This bound is dependent on the underlying statistics of the data, labels, and sensitive group attributes. The authors validate their findings through empirical analysis on three real-world datasets: COMPAS, Adult, and Law School. They compare their upper bound to existing fair classifiers in the literature, showing that achieving high accuracy with low bias could be fundamentally limited based on statistical disparities across groups. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Machine learning is being used more often in important decisions like hiring, law enforcement, and finance. But it’s crucial that these systems are fair for everyone. This paper looks at how to balance fairness and accuracy in machine learning models. The authors come up with a new way to measure the tradeoff between fairness and accuracy. They test their idea on three real-world datasets and compare it to other approaches. |
Keywords
» Artificial intelligence » Machine learning
