Inside a predator-prey model for multi-objective optimization

Christian Grimme, Karlheinz Schmitt
2006 Proceedings of the 8th annual conference on Genetic and evolutionary computation - GECCO '06  
In this article, new variation operators for evolutionary multiobjective algorithms (EMOA) are proposed. On the basis of a predator-prey model theoretical considerations as well as empirical results lead to the development of a new recombination operator, which improves the approximation of the set of efficient solutions significantly. Furtheron, it is shown that applying speciation to the analysed model makes it possible to handle even more complex problems.
doi:10.1145/1143997.1144121 dblp:conf/gecco/GrimmeS06 fatcat:w5dzm4hx4vff3bieg7t7gsgvni