Multivariate concentration of measure type results using exchangeable pairs and size biasing

TL;DR

This paper develops new multivariate concentration of measure inequalities using exchangeable pairs and size biasing, with applications to U-statistics and permutation statistics in nonparametric testing.

Contribution

It introduces a novel framework for multivariate concentration inequalities based on exchangeable pairs, extending existing results to cases with non-zero remainder terms.

Findings

Derived concentration inequalities for exchangeable pairs with zero remainder.

Extended results to permutation statistics like Mann-Whitney-Wilcoxon.

Provided applications to nonparametric statistical tests.

Abstract

Let $(\mathbf{W,W'})$ be an exchangeable pair of vectors in $\mathbb{R}^k$. Suppose this pair satisfies \beas E(\mathbf{W}'|\mathbf{W})=(I_k-\Lambda)\mathbf{W}+\mathbf{R(W)}. \enas If $||\mathbf{W-W'}||_2\le K$ and $\mathbf{R(W)}=0$, then concentration of measure results of following form is proved for all $\mathbf{w}\succeq 0$ when the moment generating function of $\mathbf{W}$ is finite. \beas P(\mathbf{W}\succeq\mathbf{w}),P(\mathbf{W}\preceq -\mathbf{w})\le \exp(-\frac{||\mathbf{w}||_2^2}{2K^2\nu_1}), \enas for an explicit constant $\nu_1$, where $\succeq$ stands for coordinate wise $\ge$ ordering. This result is applied to examples like complete non degenerate U-statistics. Also, we deal with the example of doubly indexed permutation statistics where $\mathbf{R(W)}\neq 0$ and obtain similar concentration of measure inequalities. Practical examples from doubly indexed permutation statistics include Mann-Whitney-Wilcoxon statistic and random intersection of two graphs. Both these two examples are used in nonparametric statistical testing. We conclude the paper with a multivariate generalization of a recent concentration result due to Ghosh and Goldstein \cite{cnm} involving bounded size bias couplings.