License: Creative Commons Attribution 4.0 International license (CC BY 4.0)
When quoting this document, please refer to the following
DOI: 10.4230/LIPIcs.ESA.2022.56
URN: urn:nbn:de:0030-drops-169946
URL: https://drops.dagstuhl.de/opus/volltexte/2022/16994/
Go to the corresponding LIPIcs Volume Portal


Friggstad, Zachary ; Jamshidian, Mahya

Improved Polynomial-Time Approximations for Clustering with Minimum Sum of Radii or Diameters

pdf-format:
LIPIcs-ESA-2022-56.pdf (0.7 MB)


Abstract

We give an improved approximation algorithm for two related clustering problems. In the Minimum Sum of Radii clustering problem (MSR), we are to select k balls in a metric space to cover all points while minimizing the sum of the radii of these balls. In the Minimum Sum of Diameters clustering problem (MSD), we are to simply partition the points of a metric space into k parts while minimizing the sum of the diameters of these parts. We present a 3.389-approximation for MSR and a 6.546-approximation for MSD, improving over their respective 3.504 and 7.008 approximations developed by Charikar and Panigrahy (2001). In particular, our guarantee for MSD is better than twice our guarantee for MSR.
Our approach refines a so-called bipoint rounding procedure of Charikar and Panigrahy’s algorithm by considering centering balls at some points that were not necessarily centers in the bipoint solution. This added versatility enables the analysis of our improved approximation guarantees. We also provide an alternative approach to finding the bipoint solution using a straightforward LP rounding procedure rather than a primal-dual algorithm.

BibTeX - Entry

@InProceedings{friggstad_et_al:LIPIcs.ESA.2022.56,
  author =	{Friggstad, Zachary and Jamshidian, Mahya},
  title =	{{Improved Polynomial-Time Approximations for Clustering with Minimum Sum of Radii or Diameters}},
  booktitle =	{30th Annual European Symposium on Algorithms (ESA 2022)},
  pages =	{56:1--56:14},
  series =	{Leibniz International Proceedings in Informatics (LIPIcs)},
  ISBN =	{978-3-95977-247-1},
  ISSN =	{1868-8969},
  year =	{2022},
  volume =	{244},
  editor =	{Chechik, Shiri and Navarro, Gonzalo and Rotenberg, Eva and Herman, Grzegorz},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops.dagstuhl.de/opus/volltexte/2022/16994},
  URN =		{urn:nbn:de:0030-drops-169946},
  doi =		{10.4230/LIPIcs.ESA.2022.56},
  annote =	{Keywords: Approximation Algorithms, Clustering, Linear Programming}
}

Keywords: Approximation Algorithms, Clustering, Linear Programming
Collection: 30th Annual European Symposium on Algorithms (ESA 2022)
Issue Date: 2022
Date of publication: 01.09.2022


DROPS-Home | Fulltext Search | Imprint | Privacy Published by LZI