arXiv:2602.22334cs.LGcs.AI2026-02

揭示了平衡混合中峰度对比度随维度衰减的1/R规律,解释了传统方法失效原因。

A 1/R Law for Kurtosis Contrast in Balanced Mixtures

  • 提出有效宽度R下峰度衰减的1/R理论,证明其紧致性
  • 在样本量T条件下,突破1/√T估计精度需R≤κ_max√T
  • 通过符号一致源筛选可恢复与维数无关的对比度,适合高维信号分离

基于峰度的独立成分分析在宽且平衡的混合中性能下降。我们证明了一个精确的冗余定律:对于标准化投影,其有效宽度$R_{\mathrm{eff}}$(参与度比),总体超额峰度满足$|κ(y)|=O(κ_{\max}/R_{\mathrm{eff}})$,在平衡条件下给出紧致的$O(c_bκ_{\max}/R)$界(通常$c_b=O(\log R)$)。作为不可能性判据,在样本峰度估计的标准有限矩条件下,超越$O(1/\sqrt{T})$估计尺度需满足$R\lesssim κ_{\max}\sqrt{T}$。此外,我们证明了‘净化’——选取$m\ll R$个符号一致源——可恢复与$R$无关的对比度$Ω(1/m)$,并提出简单数据驱动启发式方法。合成实验验证了预测的衰减规律、√T转折点及对比度恢复。

原文摘要 · Abstract (English)

Kurtosis-based Independent Component Analysis (ICA) weakens in wide, balanced mixtures. We prove a sharp redundancy law: for a standardized projection with effective width $R_{\mathrm{eff}}$ (participation ratio), the population excess kurtosis obeys $|κ(y)|=O(κ_{\max}/R_{\mathrm{eff}})$, yielding the order-tight $O(c_bκ_{\max}/R)$ under balance (typically $c_b=O(\log R)$). As an impossibility screen, under standard finite-moment conditions for sample kurtosis estimation, surpassing the $O(1/\sqrt{T})$ estimation scale requires $R\lesssim κ_{\max}\sqrt{T}$. We also show that \emph{purification} -- selecting $m\!\ll\!R$ sign-consistent sources -- restores $R$-independent contrast $Ω(1/m)$, with a simple data-driven heuristic. Synthetic experiments validate the predicted decay, the $\sqrt{T}$ crossover, and contrast recovery.

独立成分分析峰度分析高维统计信号分离

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。