Evolutionary algorithms may be applied in modeling in a number of intrinsically different ways. At one extreme, evolutionary algorithms are interchangeable tools; any alternative search mechanism would be equally
acceptable if it produced reasonable results. At the opposite extreme, the evolutionary nature of the algorithm is essential, the system being intended to directly simulate the evolutionary behavior of the system under study. Thus, evolutionary algorithms merge imperceptibly into direct simulations of ecosystems' evolutionary dynamics. There is a wide range of applications lying in between. We examine six classes in more detail.
Was this article helpful?