Improved approximation algorithms for two Euclidean k-Center variants
Haris Angelidakis, Ivan Sergeev, Pontus Westermark
备注:
An approximation algorithm with the same factor for Robust Euclidean k-Supplier also recently appeared in an independent work of Lee, Nagarajan and Wang (arXiv:2112.01700)
The $k$-Center problem is one of the most popular clustering problems. After decades of work, the complexity of most of its variants on general metrics is now well understood. Surprisingly, this is not the case for a natural setting that often arises in practice, namely the Euclidean setting, in which the input points are points in $\mathbb{R}^d$, and the distance between them is the standard $\ell_2$ Euclidean distance. In this work, we study two Euclidean $k$-Center variants, the Matroid Center problem on the real line and the Robust Euclidean $k$-Supplier problem, and provide algorithms that improve upon the best approximation guarantees known for these problems. In particular, we present a simple $2.5$-approximation algorithm for the Matroid Center problem on the real line, thus improving upon the $3$-approximation factor algorithm of Chen, Li, Liang, and Wang (2016) that works for general metrics. Moreover, we present a $(1 + \sqrt{3})$-approximation algorithm for the Robust Euclidean $k$-Supplier problem, thus improving upon the state-of-the-art $3$-approximation algorithm for Robust $k$-Supplier on general metrics and matching the best approximation factor known for the non-robust setting by Nagarajan, Schieber and Shachnai (2020).
Fine-Grained Complexity of Continuous Euclidean k-Center
University of Bonn(波恩大学);ETH Zurich(苏黎世联邦理工学院);Saarland University(萨尔兰大学);Max Planck Institute for Informatics(马克斯·普朗克计算机科学研究所);The University of Sheffield(谢菲尔德大学);Rutgers University(罗格斯大学);Hausdorff Center for Mathematics(豪斯多夫数学中心);Lamarr Institute for Machine Learning and Artificial Intelligence(拉玛尔机器学习与人工智能研究所);University of Massachusetts, Amherst(马萨诸塞大学阿默斯特分校);Aalto University(阿尔托大学)
在欧几里德$k$ -均值问题中,我们给出了欧几里德空间中的$n$点的集合$D$和正整数$k$。我们的目标是在同一空间(中心)中识别一组$k$点,以最小化$D$中每个点与最近中心之间的欧氏距离平方和。已知该问题是APX困难的,并且当前最佳近似比是基于该问题的标准LP的原始-对偶$6.357$近似[Ahmadian et al. FOCS'17,SICOMP'20]。在本说明中,我们将展示如何一个小的修改艾哈迈迪等。的分析导致略微改进的$6.12903$近似。作为相关结果,我们还证明了所述线性规划至少具有$\frac{16+\sqrt{5}}{15}>1.2157$的完整性间隙.
On Coresets for Fair Clustering in Metric and Euclidean Spaces and Their Applications
University of Bergen(卑尔根大学)
公平聚类是聚类的一种受约束的变体,其目标是划分一组彩色点,使得每个聚类中任何颜色的点的分数或多或少等于数据集中该颜色的点的分数。这种变体最近由Chierichetti等人[NeurIPS,2017]在一项开创性的工作中引入,并在聚类文献中广泛流行。在本文中,我们提出了一种新的构造的核心集的公平聚类的基础上随机抽样。新的结构使我们能够获得第一个coreset公平聚类一般度量空间。对于欧氏空间,我们得到了第一个其大小不指数依赖于维数的coreset。我们的coreset结果解决了Schmidt et al. [WAOA,2019]和Huang et al. [NeurIPS,2019]提出的开放问题。新的coreset结构有助于设计几个新的近似和流算法。特别地,我们得到了第一个度量公平聚类的真常数-近似算法,其运行时间是固定参数可处理的(FPT)。在欧几里德的情况下,我们推导出第一个$(1+ε)$ -近似算法的公平聚类的时间复杂度是近线性的,并不依赖于空间的维数指数。此外,我们的核心集建设计划是相当普遍的,并产生了广泛的约束聚类问题的核心集。这导致这些问题在一般度量和近线性时间$(1+ε)$ -近似的欧几里德度量的改进常数近似。