Learning the large-scale structure of the max-sat landscape using populations
Learning the large-scale structure of the max-sat landscape using populations
A new algorithm for solving MAX-SAT problems is introduced which clusters good solutions, and restarts the search from the closest feasible solution to the centroid of each cluster. This is shown to be highly efficient for finding good solutions of large MAX-SAT problems. We argue that this success is due to the population learning the large-scale structure of the fitness landscape. Systematic studies of the landscape are presented to support this hypothesis. In addition, a number of other strategies are tested to rule out other possible explanations of the success. Preliminary results are shown indicating that extensions of the proposed algorithm can give similar improvements on other hard optimisation problems.
518-529
Qasem, Mohamed
e7828af0-2b26-4142-85a6-1e92298af47c
Prugel-Bennett, Adam
b107a151-1751-4d8b-b8db-2c395ac4e14e
August 2010
Qasem, Mohamed
e7828af0-2b26-4142-85a6-1e92298af47c
Prugel-Bennett, Adam
b107a151-1751-4d8b-b8db-2c395ac4e14e
Qasem, Mohamed and Prugel-Bennett, Adam
(2010)
Learning the large-scale structure of the max-sat landscape using populations.
IEEE Transactions on Evolutionary Computation, 14 (4), .
(doi:10.1109/TEVC.2009.2033579).
Abstract
A new algorithm for solving MAX-SAT problems is introduced which clusters good solutions, and restarts the search from the closest feasible solution to the centroid of each cluster. This is shown to be highly efficient for finding good solutions of large MAX-SAT problems. We argue that this success is due to the population learning the large-scale structure of the fitness landscape. Systematic studies of the landscape are presented to support this hypothesis. In addition, a number of other strategies are tested to rule out other possible explanations of the success. Preliminary results are shown indicating that extensions of the proposed algorithm can give similar improvements on other hard optimisation problems.
Text
paper.pdf
- Author's Original
More information
e-pub ahead of print date: 15 December 2009
Published date: August 2010
Organisations:
Southampton Wireless Group
Identifiers
Local EPrints ID: 268060
URI: http://eprints.soton.ac.uk/id/eprint/268060
PURE UUID: de4c56b2-2fff-42d0-b4e7-17461f098e15
Catalogue record
Date deposited: 19 Oct 2009 08:34
Last modified: 14 Mar 2024 09:03
Export record
Altmetrics
Contributors
Author:
Mohamed Qasem
Author:
Adam Prugel-Bennett
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics