Summary of When Can Memorization Improve Fairness?, by Bob Pepin et al.
When Can Memorization Improve Fairness?
by Bob Pepin, Christian Igel, Raghavendra Selvan
First submitted to arxiv on: 12 Dec 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Computers and Society (cs.CY)
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 investigates how additive fairness metrics, specifically statistical parity, equal opportunity, and equalized odds, can be manipulated in a multi-class classification problem by memorizing a subset of the population. The authors provide explicit expressions for the resulting bias, which depends on the label and group membership distribution of the memorized dataset and the classifier’s bias on the unmemorized data. Additionally, they characterize the memorized datasets that eliminate biases for all three metrics considered. Furthermore, upper and lower bounds are given for the total probability mass in the memorized dataset necessary for complete elimination of these biases. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper looks at how to make a machine learning model be fairer by looking at groups of people and giving them different results. The authors show that making the model “remember” certain people can make it more unfair, but they also give rules for when this won’t happen. They even find limits on how many people need to be remembered to make the model completely fair. |
Keywords
» Artificial intelligence » Classification » Machine learning » Probability
