Summary of Fairness Without Sensitive Attributes Via Knowledge Sharing, by Hongliang Ni et al.
Fairness without Sensitive Attributes via Knowledge Sharing
by Hongliang Ni, Lei Han, Tong Chen, Shazia Sadiq, Gianluca Demartini
First submitted to arxiv on: 27 Sep 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 This paper proposes a novel approach to improving model fairness, addressing the challenge of missing sensitive attribute values due to growing concerns about data privacy. The authors introduce “Reckoner”, a confidence-based hierarchical classifier structure that reliably learns fair models even when sensitive demographic information is unavailable. The key innovation is a dual-model system where one model learns from high-confidence data and another from low-confidence data, allowing it to avoid biased predictions. Experimental results show that Reckoner outperforms state-of-the-art baselines in terms of both accuracy and fairness metrics on the COMPAS and New Adult datasets. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about making sure AI models don’t make unfair decisions. Right now, people are worried about keeping their personal information private, so we can’t always use sensitive details like age or gender to train these models. The authors came up with a clever idea called “Reckoner” that helps the model learn to be fair even without this information. It works by having two separate models: one learns from very confident predictions and another from less confident ones. This way, Reckoner can avoid making unfair decisions and do better than other methods in terms of being both accurate and fair. |
