中文

高维多变量 Kendall-$\tau$ 的极限谱分布

统计理论 2025-11-25 v3 概率论 统计理论

摘要

多变量 Kendall-τ\tau 统计量记为 KnK_n, 在稳健统计分析中发挥重要作用。本文建立了 KnK_n 经验谱分布(ESD)的极限性质。我们证明了 12pKn\frac{1}{2}pK_n 的 ESD几乎必然收敛于 Mar\v{c}enko--Pastur 定律,其方差参数为 12\frac{1}{2}, 这类似于经典样本协方差矩阵的结果。Using Stieltjes transform techniques, we extend these results to the independent component model, deriving a fixed-point equation that characterizes the limiting spectral distribution of 12trΣKn\frac{1}{2}tr\Sigma K_n. The theoretical findings are validated through comprehensive simulation studies.

关键词

引用

@article{arxiv.2510.21077,
  title  = {Limiting Spectral Distribution of High-dimensional Multivariate Kendall-$\tau$},
  author = {Ruoyu Wu},
  journal= {arXiv preprint arXiv:2510.21077},
  year   = {2025}
}