arXiv:2510.07746cs.LG2025-10被引 2
t-SNE会夸大聚类效果，结果不可靠。
t-SNE Exaggerates Clusters, Provably
- 证明t-SNE会放大输入数据的聚类强度
- 异常点越极端，可视化中聚类越明显
- 提醒用户谨慎解读t-SNE结果，尤其在聚类分析时
t-SNE广泛用于数据可视化，人们普遍认为其输出结构与原始数据一致。然而我们证明：(1) 输入数据的聚类强度无法从t-SNE结果中可靠推断；(2) 异常点越极端，其在可视化中形成的聚类越显著。我们在实际数据中验证了这些失效模式的普遍存在，警示用户在依赖t-SNE进行聚类分析时需保持警惕。
原文摘要 · Abstract (English)
Central to the widespread use of t-distributed stochastic neighbor embedding (t-SNE) is the conviction that it produces visualizations whose structure roughly matches that of the input. To the contrary, we prove that (1) the strength of the input clustering, and (2) the extremity of outlier points, cannot be reliably inferred from the t-SNE output. We demonstrate the prevalence of these failure modes in practice as well.
降维可视化聚类
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。