Summary of Learning For Counterfactual Fairness From Observational Data, by Jing Ma et al.
Learning for Counterfactual Fairness from Observational Data
by Jing Ma, Ruocheng Guo, Aidong Zhang, Jundong Li
First submitted to arxiv on: 17 Jul 2023
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Computers and Society (cs.CY); Machine Learning (stat.ML)
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 framework called CLAIRE for achieving counterfactually fair predictions from observational data without prior knowledge of the underlying causal model. The authors address the challenge of mitigating biases from sensitive attributes while maintaining high prediction performance. They introduce a representation learning framework based on counterfactual data augmentation and an invariant penalty, which effectively reduces bias under certain general assumptions. CLAIRE is evaluated on both synthetic and real-world datasets, demonstrating its superiority in terms of both counterfactual fairness and prediction performance. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps us make sure that machines are fair when making decisions about people. Right now, there are ways to make machines more fair, but they need to know what the underlying rules are for how things work. The problem is that these rules can be hard to figure out, and if we get them wrong, it could lead to unfair results. This paper introduces a new way called CLAIRE that allows us to make predictions without knowing the underlying rules. It does this by using data and special penalties to help reduce bias. The results show that CLAIRE is better than other methods at being fair and making good predictions. |
Keywords
* Artificial intelligence * Data augmentation * Representation learning
