Summary of Addressing Polarization and Unfairness in Performative Prediction, by Kun Jin et al.
Addressing Polarization and Unfairness in Performative Prediction
by Kun Jin, Tian Xie, Yang Liu, Xueru Zhang
First submitted to arxiv on: 24 Jun 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); 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 explores the societal implications of performative prediction (PP) solutions, which explicitly consider model-dependent distribution shifts when learning machine learning models. The authors examine the fairness property of PS solutions in PP, finding that they can incur severe polarization effects and group-wise loss disparity. Existing fairness mechanisms may mitigate unfairness but fail to achieve stability under model-dependent distribution shifts. To address this, the authors propose novel fairness intervention mechanisms that simultaneously achieve stability and fairness in PP settings. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Performative prediction is a framework for machine learning models used in applications involving humans. When these models are used, they can change the way data is distributed. This paper looks at how performative prediction solutions affect fairness. The results show that these solutions can cause problems like polarization and unfair treatment of certain groups. The authors also explore existing methods to fix these issues but find that they don’t work well when there are changes in the data distribution. To solve this, the paper proposes new ways to make sure both stability and fairness are achieved. |
Keywords
» Artificial intelligence » Machine learning
