Repository logo

Infoscience

  • English
  • French
Log In
Logo EPFL, École polytechnique fédérale de Lausanne

Infoscience

  • English
  • French
Log In
  1. Home
  2. Academic and Research Output
  3. Journal articles
  4. Learning from survey propagation: a neural network for MAX-E-3-SAT
 
research article

Learning from survey propagation: a neural network for MAX-E-3-SAT

Marino, Raffaele  
September 1, 2021
Machine Learning-Science And Technology

Many natural optimization problems are NP-hard, which implies that they are probably hard to solve exactly in the worst-case. However, it suffices to get reasonably good solutions for all (or even most) instances in practice. This paper presents a new algorithm for computing approximate solutions in Theta(N) for the maximum exact 3-satisfiability (MAX-E-3-SAT) problem by using supervised learning methodology. This methodology allows us to create a learning algorithm able to fix Boolean variables by using local information obtained by the Survey Propagation algorithm. By performing an accurate analysis, on random conjunctive normal form instances of the MAX-E-3-SAT with several Boolean variables, we show that this new algorithm, avoiding any decimation strategy, can build assignments better than a random one, even if the convergence of the messages is not found. Although this algorithm is not competitive with state-of-the-art maximum satisfiability solvers, it can solve substantially larger and more complicated problems than it ever saw during training.

  • Details
  • Metrics
Type
research article
DOI
10.1088/2632-2153/ac0496
Web of Science ID

WOS:000674928900001

Author(s)
Marino, Raffaele  
Date Issued

2021-09-01

Publisher

IOP PUBLISHING LTD

Published in
Machine Learning-Science And Technology
Volume

2

Issue

3

Article Number

035032

Subjects

Computer Science, Artificial Intelligence

•

Computer Science, Interdisciplinary Applications

•

Multidisciplinary Sciences

•

Computer Science

•

Science & Technology - Other Topics

•

maximum satisfiability

•

message passing

•

combinatorial optimization

•

deep learning

•

sat

•

strategies

•

algorithms

•

breaking

Editorial or Peer reviewed

REVIEWED

Written at

EPFL

EPFL units
LTHC  
Available on Infoscience
August 14, 2021
Use this identifier to reference this record
https://infoscience.epfl.ch/handle/20.500.14299/180575
Logo EPFL, École polytechnique fédérale de Lausanne
  • Contact
  • infoscience@epfl.ch

  • Follow us on Facebook
  • Follow us on Instagram
  • Follow us on LinkedIn
  • Follow us on X
  • Follow us on Youtube
AccessibilityLegal noticePrivacy policyCookie settingsEnd User AgreementGet helpFeedback

Infoscience is a service managed and provided by the Library and IT Services of EPFL. © EPFL, tous droits réservés