Discovery - University of Dundee - Online Publications

Library & Learning Centre

Self-learning genetic algorithm for constrains satisfaction problems

Self-learning genetic algorithm for constrains satisfaction problems

Research output: Chapter in Book/Report/Conference proceedingChapter

View graph of relations

Authors

Research units

Info

Original languageEnglish
Title of host publication2012 Imperial College Computing Student Workshop
Subtitle of host publicationICCSW 2012
EditorsAndrew V. Jones
Place of PublicationLeibnitz, Austria
PublisherDagstuhl Publications
Pages156-162
Number of pages7
Volume28
ISBN (Print)9783939897484
DOIs
StatePublished - 2012
Event2nd Imperial College Computing Student Workshop, ICCSW 2012 - London, United Kingdom

Publication series

NameOpen Access Series in Informatics (OASIcs)
PublisherDagstuhl
Volume28
ISSN (Print)2190-6807

Conference

Conference2nd Imperial College Computing Student Workshop, ICCSW 2012
CountryUnited Kingdom
CityLondon
Period27/09/1228/09/12
Internet address

Abstract

The efficient choice of a preprocessing level can reduce the search time of a constraint solver to find a solution to a constraint problem. Currently the parameters in constraint solver are often picked by hand by experts in the field. Genetic algorithms are a robust machine learning technology for problem optimization such as function optimization. Self-learning Genetic Algorithm are a strategy which suggests or predicts the suitable preprocessing method for large scale problems by learning from the same class of small scale problems. In this paper Self-learning Genetic Algorithms are used to create an automatic preprocessing selection mechanism for solving various constraint problems. The experiments in the paper are a proof of concept for the idea of combining genetic algorithm self-learning ability with constraint programming to aid in the parameter selection issue. © Hu Xu and Karen Petrie.

Documents

DOI

Library & Learning Centre

Contact | Accessibility | Policy