Abstract:

Classical evolutionary algorithms have been extremely successful at solving certain problems. But they implement a very simple model of evolutionary biology that misses out several aspects that might be exploited by more sophisticated algorithms. We have previously critiqued the traditional naïve approach to bio-inspired algorithm design, that moves straight from a simplistic description of the biology into some algorithm. Here we present a process for developing richer evolutionary algorithms abstracted from various processes of biological evolution, with a corresponding richer analogical computational structure, and indicate how that might be further abstracted.