无监督表示会意外泄露年龄收入等敏感信息，打破公平假设。
SOMtime the World Ain$'$t Fair: Violating Fairness Using Self-Organizing Maps
- 用自组织映射构建拓扑保持的嵌入，隐含敏感属性
- 在真实数据上相关性达0.85，远超传统方法
- 揭示无监督模型也存在公平风险，需审计表示层
无监督表示通常被认为在隐藏敏感属性时是中立的。我们证明这一假设不成立。通过基于高容量自组织映射的SOMtime方法，我们发现年龄、收入等敏感属性在纯无监督嵌入中仍作为主导潜在轴出现，即使这些属性未被输入。在世界价值观调查（五个国家）和Census-Income数据集上，SOMtime恢复了与隐藏敏感属性对齐的单调排序，斯皮尔曼相关系数最高达0.85；而PCA和UMAP通常低于0.23（单一例外为0.31），t-SNE和自编码器最多达到0.34。此外，SOMtime嵌入的无监督聚类产生人口统计学偏斜的簇，表明即使无监督任务也存在下游公平风险。这些发现表明，对有序敏感属性而言，‘因无知而公平’在表示层面失效，公平审计必须扩展至机器学习流水线中的无监督组件。代码已开源。
原文摘要 · Abstract (English)
Unsupervised representations are widely assumed to be neutral with respect to sensitive attributes when those attributes are withheld from training. We show that this assumption is false. Using SOMtime, a topology-preserving representation method based on high-capacity Self-Organizing Maps, we demonstrate that sensitive attributes such as age and income emerge as dominant latent axes in purely unsupervised embeddings, even when explicitly excluded from the input. On two large-scale real-world datasets (the World Values Survey across five countries and the Census-Income dataset), SOMtime recovers monotonic orderings aligned with withheld sensitive attributes, achieving Spearman correlations of up to 0.85, whereas PCA and UMAP typically remain below 0.23 (with a single exception reaching 0.31), and against t-SNE and autoencoders which achieve at most 0.34. Furthermore, unsupervised segmentation of SOMtime embeddings produces demographically skewed clusters, demonstrating downstream fairness risks without any supervised task. These findings establish that \textit{fairness through unawareness} fails at the representation level for ordinal sensitive attributes and that fairness auditing must extend to unsupervised components of machine learning pipelines. We have made the code available at~ https://github.com/JosephBingham/SOMtime
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。