Evolutionary stable and unstable strategies in neural networks

Neural networks simulate simple organisms that evolve in a two-dimensional environment (with selective reproduction and mutation) on the basis of a fitness criterion which rewards residing in favorable environmental zones and obtaining food. The evolution of two different populations of organisms is studied: organisms that are informed of the environmental zone they are in at any particular time and organisms that are not so informed. The results show that the behavioral strategies that evolve in the two populations closely reflect the differences between the two types of organisms. Furthermore, the course of evolution can be divided into different successive phases which are characterized by different behavioral strategies. The initial strategies are evolutionarily unstable because they tend to be replaced by new more effective strategies introduced by mutants, until final stable strategies emerge which differ in the two populations because of their different evolutionary history.<
>