Operator Choice and the Evolution of Robust Solutions

Abstract

This research demonstrates that evolutionary pressurefavouring robust solutions has a significant impact onthe evolutionary process. More robust solutions aresolutions that are less likely to be degraded by thegenetic operators. This pressure for robust solutionscan be used to explain a number of evolutionarybehaviours. The experiments examine the effect ofdifferent types and rates of genetic operators on theevolution of robust solutions. Previously robustnesswas observed to occur through an increase ininoperative genes (introns). This work shows thatalternative strategies to increase robustness canevolve. The results also show that different geneticoperators lead to different strategies for improvingrobustness. These results can be useful in designinggenetic operators to encourage particular evolutionarybehaviors.