DMCA
Advanced Scatter Search for the Max-Cut Problem
| Citations: | 13 - 2 self |
BibTeX
@MISC{Martí_advancedscatter,
author = {Rafael Martí and Abraham Duarte and Juan Carlos and Manuel Laguna},
title = {Advanced Scatter Search for the Max-Cut Problem},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
The Max-Cut problem consists of finding a partition of the nodes of a weighted graph into two subsets such that the sum of the weights between both sets is maximized. This is an NP-hard problem that can also be formulated as an integer quadratic program. Several solution methods have been developed since the 1970s and applied to a variety of fields, particularly in engineering and layout design. We propose a heuristic method — based on the scatter search methodology — for finding approximate solutions to this optimization problem. Our solution procedure incorporates some innovative features within the scatter search framework: 1) the solution of the maximum diversity problem to increase diversity in the reference set, 2) a dynamic adjustment of a key parameter within the search, and 3) the adaptive selection of a combination method. We perform extensive computational experiments to first study the effect of changes in critical scatter search elements and then to compare the efficiency of our proposal with previous solution procedures. Keywords: Max-cut problem, scatter search, metaheuristics, evolutionary algorithms Version: June 1, 2007Advanced Scatter Search for the Max-Cut Problem / 2 1.
Keyphrases
max-cut problem advanced scatter search scatter search previous solution procedure key parameter several solution method scatter search framework integer quadratic program layout design combination method evolutionary algorithm version critical scatter search element optimization problem maximum diversity problem extensive computational experiment scatter search methodology dynamic adjustment weighted graph np-hard problem approximate solution innovative feature solution procedure reference set heuristic method adaptive selection
