Measuring Association on Topological Spaces Using Kernels and Geometric Graphs

Measuring Association on Topological Spaces Using Kernels and Geometric Graphs
复制标题

使用核和几何图测量拓扑空间上的关联

DOI:
--
复制
发表时间:
2020
期刊:
影响因子:
--
通讯作者:
B. Sen
B. Sen
中科院分区:
--
文献类型:
--
作者:
Nabarun Deb;Promit Ghosal;B. Sen

文献摘要

参考文献

被引文献

相似文献

在本文中,我们提出并研究了一类简单的,非参数的,但可解释的措施之间的关联两个随机变量$X$和$Y$采取的值在一般拓扑空间。这些非参数测度--使用再生核希尔伯特空间的理论定义--捕获了X和Y之间的依赖强度,并且具有这样的性质:当且仅当变量是独立的时,它们为0;当且仅当一个变量是另一个变量的可测函数时,它们为1。此外,这些人口的措施,可以一致地估计使用的一般框架图泛函,其中包括$k$-最近邻图和最小生成树。此外,这些估计的一个子类也显示,以适应潜在的分布的内在维度。这些经验测量中的一些也可以在近似线性时间内计算。在X$和Y$之间独立的假设下,这些经验测量(适当归一化)具有标准正态极限分布。因此,这些措施也可以很容易地用来检验假设的相互独立性之间的X$和$Y$。事实上,据我们所知,这些是唯一拥有所有上述理想特性的程序。此外,当限制到欧氏空间,我们可以使这些样本测量的关联有限样本分布自由,在独立的假设下,通过使用多元秩定义的最优运输理论。最近在Dette et al.(2013),Chatterjee(2019)和Azadkia and Chatterjee(2019)中提出的相关系数可以被视为这类一般度量的特例。
In this paper we propose and study a class of simple, nonparametric, yet interpretable measures of association between two random variables $X$ and $Y$ taking values in general topological spaces. These nonparametric measures -- defined using the theory of reproducing kernel Hilbert spaces -- capture the strength of dependence between $X$ and $Y$ and have the property that they are 0 if and only if the variables are independent and 1 if and only if one variable is a measurable function of the other. Further, these population measures can be consistently estimated using the general framework of graph functionals which include $k$-nearest neighbor graphs and minimum spanning trees. Moreover, a sub-class of these estimators are also shown to adapt to the intrinsic dimensionality of the underlying distribution. Some of these empirical measures can also be computed in near linear time. Under the hypothesis of independence between $X$ and $Y$, these empirical measures (properly normalized) have a standard normal limiting distribution. Thus, these measures can also be readily used to test the hypothesis of mutual independence between $X$ and $Y$. In fact, as far as we are aware, these are the only procedures that possess all the above mentioned desirable properties. Furthermore, when restricting to Euclidean spaces, we can make these sample measures of association finite-sample distribution-free, under the hypothesis of independence, by using multivariate ranks defined via the theory of optimal transport. The recent correlation coefficient proposed in Dette et al. (2013), Chatterjee (2019), and Azadkia and Chatterjee (2019) can be seen as a special case of this general class of measures.
DOI: 10.13140/rg.2.2.27112.37120
发表时间: 2017-08
期刊: J. Mach. Learn. Res.
影响因子: --
作者:
Z. Szabó;Bharath K. Sriperumbudur
通讯作者: Z. Szabó;Bharath K. Sriperumbudur
DOI: --
发表时间: 2015-05
期刊: ArXiv
影响因子: --
作者:
Yakir A. Reshef;David N. Reshef;H. Finucane;Pardis C Sabeti;M. Mitzenmacher
通讯作者: Yakir A. Reshef;David N. Reshef;H. Finucane;Pardis C Sabeti;M. Mitzenmacher
DOI: --
发表时间: 2019-06
期刊: arXiv: Methodology
影响因子: --
作者:
Divyansh Agarwal;Somabha Mukherjee;B. Bhattacharya;N. Zhang
通讯作者: Divyansh Agarwal;Somabha Mukherjee;B. Bhattacharya;N. Zhang
DOI: 10.1080/01621459.2021.1923508
发表时间: 2021-06-16
影响因子: 3.7
作者:
Deb,Nabarun;Sen,Bodhisattva
通讯作者: Sen,Bodhisattva
数据科学中的随机近邻图、巨型组件和应用。
DOI: 10.1017/jpr.2020.21
发表时间: 2020
影响因子: 1
作者:
Linderman,GeorgeC;Mishne,Gal;Jaffe,Ariel;Kluger,Yuval;Steinerberger,Stefan
通讯作者: Steinerberger,Stefan