TUHH Open Research
Help
  • Log In
    New user? Click here to register.Have you forgotten your password?
  • English
  • Deutsch
  • Communities & Collections
  • Publications
  • Research Data
  • People
  • Institutions
  • Projects
  • Statistics
  1. Home
  2. TUHH
  3. Publications
  4. noSAT-MaxSATv2
 
Options

noSAT-MaxSATv2

Citation Link: https://doi.org/10.15480/882.9031
Publikationstyp
Conference Paper
Date Issued
2023
Sprache
English
Author(s)
Lübke, Ole  orcid-logo
Softwaresysteme E-16  
Schupp, Sibylle  
Softwaresysteme E-16  
TORE-DOI
10.15480/882.9031
TORE-URI
https://hdl.handle.net/11420/44979
Start Page
27
End Page
28
Citation
MaxSAT Evaluation (2023)
Contribution to Conference
18th MaxSAT Evaluation, MSE 2023  
Publisher
Department of Computer Science, University of Helsinki
Peer Reviewed
false
Is Part Of
10138/564026
Using a SAT solver to solve MaxSAT is a well-established and efficient method. However, in resource-constrained computing environments, e.g., embedded systems, such algorithm designs can be hard to realize. With noSAT-MaxSAT, we explore alternatives that do not rely on an external, dedicated SAT solver, by executing an anytime local-search MaxSAT solving algorithm only on the hard clauses of the formula instead. In many real-world problems a significant portion of the hard clauses feature not more than two literals. Therefore, in noSAT-MaxSATv2 we integrated a linear-time algorithm for 2-SAT as a preprocessing step to find a good initial assignment before proceeding to solve the remaining hard clauses with local search. Additionally, compared to the previous version of noSAT-MaxSAT, we updated the local-search algorithm from SATLike to NuWLS, the algorithm which won all incomplete tracks of the MaxSAT Evaluation 2022.
DDC Class
004: Computer Sciences
Publication version
publishedVersion
Lizenz
https://creativecommons.org/licenses/by-sa/4.0/
Loading...
Thumbnail Image
Name

noSAT-MaxSAT.pdf

Type

Main Article

Size

105.68 KB

Format

Adobe PDF

TUHH
Weiterführende Links
  • Contact
  • Send Feedback
  • Cookie settings
  • Privacy policy
  • Impress
DSpace Software

Built with DSpace-CRIS software - Extension maintained and optimized by 4Science
Design by effective webwork GmbH

  • Deutsche NationalbibliothekDeutsche Nationalbibliothek
  • ORCiD Member OrganizationORCiD Member Organization
  • DataCiteDataCite
  • Re3DataRe3Data
  • OpenDOAROpenDOAR
  • OpenAireOpenAire
  • BASE Bielefeld Academic Search EngineBASE Bielefeld Academic Search Engine
Feedback