Summary of The Double-edged Sword Of Input Perturbations to Robust Accurate Fairness, by Xuran Li et al.
The Double-Edged Sword of Input Perturbations to Robust Accurate Fairness
by Xuran Li, Peng Wu, Yanting Chen, Xingjun Ma, Zhen Zhang, Kaixiang Dong
First submitted to arxiv on: 1 Apr 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 Medium Difficulty summary: This paper introduces a novel concept called “robust accurate fairness” that evaluates how deep neural networks (DNNs) respond to adversarial input perturbations. The authors propose an attack method called RAFair to identify biases and inaccuracies in DNNs, which can either compromise accuracy or individual fairness. They demonstrate that these biased instances can be corrected using carefully designed benign perturbations, resulting in accurate and fair predictions. This work sheds light on the double-edged sword of input perturbations and explores their impact on robust accurate fairness in DNNs. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Low Difficulty summary: This paper is about how artificial intelligence models called deep neural networks (DNNs) can be tricked into making mistakes or being unfair when faced with special types of changes to the data they’re processing. The authors have developed a way to identify these problems and fix them by using gentle, helpful changes instead. They want people to understand that AI is not perfect and needs to be improved to work fairly and accurately. |
