Online disjoint set cover without prior knowledge

Yuval Emek, Adam Goldbraikh, Erez Kantor

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

The disjoint set cover (DSC) problem is a fundamental combinatorial optimization problem concerned with partitioning the (hyper)edges of a hypergraph into (pairwise disjoint) clusters so that the number of clusters that cover all nodes is maximized. In its online version, the edges arrive one-by-one and should be assigned to clusters in an irrevocable fashion without knowing the future edges. This paper investigates the competitiveness of online DSC algorithms. Specifically, we develop the first (randomized) online DSC algorithm that guarantees a poly-logarithmic (O(log2 n)) competitive ratio without prior knowledge of the hypergraph’s minimum degree. On the negative side, we prove that the competitive ratio of any randomized online DSC algorithm must be at least (formula presented.) (even if the online algorithm does know the minimum degree in advance), thus establishing the first lower bound on the competitive ratio of randomized online DSC algorithms.

Original languageEnglish
Title of host publication27th Annual European Symposium on Algorithms, ESA 2019
EditorsMichael A. Bender, Ola Svensson, Grzegorz Herman
ISBN (Electronic)9783959771245
DOIs
StatePublished - Sep 2019
Event27th Annual European Symposium on Algorithms, ESA 2019 - Munich/Garching, Germany
Duration: 9 Sep 201911 Sep 2019

Publication series

NameLeibniz International Proceedings in Informatics, LIPIcs
Volume144
ISSN (Print)1868-8969

Conference

Conference27th Annual European Symposium on Algorithms, ESA 2019
Country/TerritoryGermany
CityMunich/Garching
Period9/09/1911/09/19

Keywords

  • Competitive analysis
  • Competitiveness with high probability
  • Disjoint set cover
  • Online algorithms

ASJC Scopus subject areas

  • Software

Fingerprint

Dive into the research topics of 'Online disjoint set cover without prior knowledge'. Together they form a unique fingerprint.

Cite this