Evolutionary Computation

In the last post I spoke of evolutionary computation. I gave some hint to the environment trying to mark it should have some role, but before describing the environment some more information about evolutionary computation must be provided. I will try to give a brief description.
Evolution implies the existence of more generations of "objects". The objects of the same generation by means of two operators,

• recombination [see picture]
• mutation

create a set of new "objects": the offspring. By means of the operator of "offspring-selection&replacement" applied to both offspring and parents a new generation is obtained.
Evolutionary programming exploits evolution to make evolve programs. I apologize for this raw description and for the lack of hint to genetic programming: I want to be as simple as possible and I cannot write everything about everything ... ^_^".
In general, evolutionary programming is managed with a top-down approach: a main program create an environment ( ... hmm interesting) and manages the generations (in this case the objects are "programs") sequentially !

Look @ Darwin if you want write some simple evolutionary application (note: I have not written evolutionary programming).

A different approach could be to exploit an environment where evolution is part or a characteristics of it and becomes an operator which subsets of involved objects can independently apply to themselves. This approach implies an asynchronous and unregulated evolution of the entire set of objects and requires high level of distribution and good management of parallelism in the underlying sub-system: some more clues in the next posts.