arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1604.03006cs.LGcs.ITmath.ITstat.ML
收藏

固定k-近邻信息估计量的解

Demystifying Fixed k-Nearest Neighbor Information Estimators

University of Illinois at Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校)

Weihao Gao, Sewoong Oh, Pramod Viswanath

摘要:

从独立同分布估计互信息从未知的联合密度函数中抽取样本是一个具有广泛应用的基本统计问题。最流行的估计量是由Kraskov和Stögbauer和Grassberger(KSG)在2004年提出的,并且是非参数的,并且基于每个样本到其$k^{\rm th}$最近邻样本的距离,其中$k$是固定的小整数。尽管它的广泛使用(科学软件包的一部分),这个估计的理论特性在很大程度上是未开发的。在本文中,我们证明了估计是一致的,也确定了一个上界的收敛速度的偏差作为样本数的函数。我们认为,优越的性能优势的KSG估计源于一个奇怪的“相关性增强”的效果,并建立在这种直觉,以新的方式修改KSG估计,以构建一个优越的估计。作为一个副产品,我们的调查,我们获得了近紧的$\ell_2$误差的固定$k$最近邻估计的微分熵Kozachenko和Leonenko的收敛速度。

英文摘要:

Estimating mutual information from i.i.d. samples drawn from an unknown joint density function is a basic statistical problem of broad interest with multitudinous applications. The most popular estimator is one proposed by Kraskov and Stögbauer and Grassberger (KSG) in 2004, and is nonparametric and based on the distances of each sample to its $k^{\rm th}$ nearest neighboring sample, where $k$ is a fixed small integer. Despite its widespread use (part of scientific software packages), theoretical properties of this estimator have been largely unexplored. In this paper we demonstrate that the estimator is consistent and also identify an upper bound on the rate of convergence of the bias as a function of number of samples. We argue that the superior performance benefits of the KSG estimator stems from a curious "correlation boosting" effect and build on this intuition to modify the KSG estimator in novel ways to construct a superior estimator. As a byproduct of our investigations, we obtain nearly tight rates of convergence of the $\ell_2$ error of the well known fixed $k$ nearest neighbor estimator of differential entropy by Kozachenko and Leonenko.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑