Evolutionary optimization through extinction dynamics

A recent method of optimization has been presented involving macroevolutionary dynamical rules. Extinction and diversification are introduced by following a set of steps where species/solutions are wired through an antisymmetric matrix of interactions. These connections are defined in terms of the fitness landscape and allow to explore the landscape in a highly effective way. It is shown that this search method is able to outperform genetic algorithms in a wide range of conditions.