Abstract
Evolutionary algorithms are often successfully applied to hard optimization problems. However, besides rules of thumb and experience, trial and error is still the leading design technique for evolutionary algorithms. A profound theoretical foundation guiding those applications is still missing. This article outlines a networked understanding of evolutionary algorithms. As a first step towards that goal, it reviews and interrelates major theoretical results and working principles in order to give an extensive insight into the internal processing of evolutionary algorithms. This not only helps to understand success and failure of evolutionary algorithms in applications but, in addition, could lead to a theory-guided design process enrichening and relieving today's heuristic techniques.