License: Creative Commons Attribution 3.0 Unported license (CC BY 3.0)
When quoting this document, please refer to the following
DOI: 10.4230/LIPIcs.ESA.2019.52
URN: urn:nbn:de:0030-drops-111730
URL: http://dagstuhl.sunsite.rwth-aachen.de/volltexte/2019/11173/
Go to the corresponding LIPIcs Volume Portal


Gottesbüren, Lars ; Hamann, Michael ; Wagner, Dorothea

Evaluation of a Flow-Based Hypergraph Bipartitioning Algorithm

pdf-format:
LIPIcs-ESA-2019-52.pdf (1 MB)


Abstract

In this paper, we propose HyperFlowCutter, an algorithm for balanced hypergraph bipartitioning that is based on minimum S-T hyperedge cuts and maximum flows. It computes a sequence of bipartitions that optimize cut size and balance in the Pareto sense, being able to trade one for the other. HyperFlowCutter builds on the FlowCutter algorithm for partitioning graphs. We propose additional features, such as handling disconnected hypergraphs, novel methods for obtaining starting S,T pairs as well as an approach to refine a given partition with HyperFlowCutter. Our main contribution is ReBaHFC, a new algorithm which obtains an initial partition with the fast multilevel hypergraph partitioner PaToH and then improves it using HyperFlowCutter as a refinement algorithm. ReBaHFC is able to significantly improve the solution quality of PaToH at little additional running time. The solution quality is only marginally worse than that of the best-performing hypergraph partitioners KaHyPar and hMETIS, while being one order of magnitude faster. Thus ReBaHFC offers a new time-quality trade-off in the current spectrum of hypergraph partitioners. For the special case of perfectly balanced bipartitioning, only the much slower plain HyperFlowCutter yields slightly better solutions than ReBaHFC, while only PaToH is faster than ReBaHFC.

BibTeX - Entry

@InProceedings{gottesbren_et_al:LIPIcs:2019:11173,
  author =	{Lars Gottesb{\"u}ren and Michael Hamann and Dorothea Wagner},
  title =	{{Evaluation of a Flow-Based Hypergraph Bipartitioning Algorithm}},
  booktitle =	{27th Annual European Symposium on Algorithms (ESA 2019)},
  pages =	{52:1--52:17},
  series =	{Leibniz International Proceedings in Informatics (LIPIcs)},
  ISBN =	{978-3-95977-124-5},
  ISSN =	{1868-8969},
  year =	{2019},
  volume =	{144},
  editor =	{Michael A. Bender and Ola Svensson and Grzegorz Herman},
  publisher =	{Schloss Dagstuhl--Leibniz-Zentrum fuer Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{http://drops.dagstuhl.de/opus/volltexte/2019/11173},
  URN =		{urn:nbn:de:0030-drops-111730},
  doi =		{10.4230/LIPIcs.ESA.2019.52},
  annote =	{Keywords: Hypergraph Partitioning, Maximum Flows, Algorithm Engineering}
}

Keywords: Hypergraph Partitioning, Maximum Flows, Algorithm Engineering
Collection: 27th Annual European Symposium on Algorithms (ESA 2019)
Issue Date: 2019
Date of publication: 06.09.2019
Supplementary Material: Source code at https://github.com/kit-algo/HyperFlowCutter


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