TY - GEN
T1 - Intensification strategies for extremal optimisation
AU - Randall, Marcus
AU - Lewis, Andrew
PY - 2010
Y1 - 2010
N2 - It is only relatively recently that extremal optimisation (EO) has been applied to combinatorial optimisation problems. As such, there have been only a few attempts to extend the paradigm to include standard search mechanisms that are routinely used by other techniques such as genetic algorithms, tabu search and ant colony optimisation. The key way to begin this process is to augment EO with attributes that it naturally lacks. While EO does not get confounded by local optima and is able to move through search space unencumbered, one of the major issues is to provide it with better search intensification strategies. In this paper, two strategies that compliment EO's mechanics are introduced and are used to augment an existing solver framework. Results, for single and population versions of the algorithm, demonstrate that intensification aids the performance of EO.
AB - It is only relatively recently that extremal optimisation (EO) has been applied to combinatorial optimisation problems. As such, there have been only a few attempts to extend the paradigm to include standard search mechanisms that are routinely used by other techniques such as genetic algorithms, tabu search and ant colony optimisation. The key way to begin this process is to augment EO with attributes that it naturally lacks. While EO does not get confounded by local optima and is able to move through search space unencumbered, one of the major issues is to provide it with better search intensification strategies. In this paper, two strategies that compliment EO's mechanics are introduced and are used to augment an existing solver framework. Results, for single and population versions of the algorithm, demonstrate that intensification aids the performance of EO.
UR - http://www.scopus.com/inward/record.url?scp=78650742187&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-17298-4_12
DO - 10.1007/978-3-642-17298-4_12
M3 - Conference contribution
AN - SCOPUS:78650742187
SN - 3642172970
SN - 9783642172977
VL - 6457 LNCS
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 115
EP - 124
BT - Simulated Evolution and Learning - 8th International Conference, SEAL 2010, Proceedings
T2 - 8th International Conference on Simulated Evolution and Learning, SEAL 2010
Y2 - 1 December 2010 through 4 December 2010
ER -
