Nearly all practical applications of Evolutionary Algorithms are hybrids and incorporate some kind of application-specific knowledge in the evolutionary procedure, resulting in a problem-specific solution. This paper focuses on two questions: How can we hybridise such that the resulting algorithm remains generally applicable and what is the best type of hybridisation? The results of detailed experiments show promising results: Evaluations could be reduced by a factor of up to 100 at best.
Print ISSN: 0178-2312
Volume: 53, 06/2005
Pages: 251 - 260