This paper investigates limiting spectral distribution of a high-dimensional Kendall's Rank correlation matrix. The underlying population is allowed to have general dependence structure. It no longer follows the Marcenko-Pastur law. It's the first result on rank correlation matrices without any independence or moment constraints
翻译:本文调查限制高维肯德尔的级相关矩阵的光谱分布。 基础人口允许有一般依赖结构。 它不再遵循马尔肯科- 帕斯图尔法律。 这是没有独立性或时间限制的级相关矩阵的第一个结果。