BriefGPT.xyz
May, 2019
缺失数据下 PCA 的相变：降低信噪比而非样本量！
Phase transition in PCA with missing data: Reduced signal-to-noise ratio, not sample size!
HTML
PDF
Niels Bruun Ipsen, Lars Kai Hansen
TL;DR
本文研究探讨缺失数据如何影响我们学习信号结构，提出了概率主成分分析的方法用于估算缺失数据模型的信号结构，理论表明缺失数据会有效降低信噪比，而不是像通常认为的降低样本量，预测出学习曲线中的临界状态和相变现象，这一结论在模拟数据和真实数据中都得到证实。
Abstract
How does
missing data
affect our ability to learn
signal structures
? It has been shown that learning signal structure in terms of principal components is dependent on the ratio of sample size and dimensionality a
→