Evolutionary computation and complex networks have received considerable
attention in recent years. Inspired by natural selection and evolutionary theory,
evolutionary algorithms (EAs) are optimization heuristics designed to solve optimization problems. They are different from classic optimization algorithms in their reliance on a population of solutions instead of searching using a single solution at a time and their ability to solve nondifferentiable functions with an arbitrary level of
complexity. They have been very successful in solving various engineering applications.
Complex networks abstract a wide range of biological and social systems
including the Internet, e-mail interactions, gene regulatory networks, science collaboration networks, phone call networks, and citation networks. Because of their significant contributions to our understanding of complex systems, complex networks have been attracting much interest and seen significant advances over the last decade.
Most of the work in the field of complex networks has concentrated on two
aspects. On the one hand, research has focused on analyzing properties found in
real-world networks, and then designing algorithms to build understanding of the
mechanisms underlying the emergence of these properties; that is to say, to design
methods that can generate the networks with certain properties. This process is
modeled as a form of complex optimization problems. On the other hand, considering the fact that EAs rely on a population of solutions, namely a set of individual solutions, the performance of EAs heavily depends on the interaction among these individuals. Complex network analysis could in principle represent this interaction
to understand evolutionary dynamics and how fitter solutions emerge.