Summary of Post-processing Fairness with Minimal Changes, by Federico Di Gennaro et al.
Post-processing fairness with minimal changes
by Federico Di Gennaro, Thibault Laugel, Vincent Grari, Xavier Renard, Marcin Detyniecki
First submitted to arxiv on: 27 Aug 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI)
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 proposes a novel post-processing algorithm for debiasing predictions that is both model-agnostic and doesn’t require sensitive attributes at test time. The algorithm aims to enforce minimal changes between biased and debiased predictions, which is a desirable but often overlooked property in fairness literature. The approach applies a multiplicative factor to the logit value of probability scores produced by a black-box classifier. Empirical evaluations on two widely used datasets demonstrate its efficacy compared to four other debiasing algorithms. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps make sure that predictions are fair and equal for everyone. It introduces a new way to do this without needing special information about the person being predicted for. The method works by changing the way probability scores are calculated, making it harder to tell if someone is biased or not. The authors tested their method on two big datasets and showed it worked better than four other ways to debias predictions. |
Keywords
» Artificial intelligence » Probability
