Thomas Bartz-Beielstein

Evolutionary Algorithms For Optimization Practitioners

Taking advantage of valuable preparational work done in the early 1960s [Bre62], these general purpose algorithms unfolded during the 1960s and 1970s. Nevertheless, optimization practitioners nowadays still experience some difficulties when applying EAs to real-world optimization problems. Our approach is problem oriented and should give optimization practitioners a working knowledge of evolutionary algorithms. Similar to the approach presented in [Kle87], we will take a look at evolutionary algorithms from the viewpoint of an optimization practitioner. We will demonstrate how problem specific knowledge can be integrated into genetic operators, and how coding and hybridization with traditional gradient search procedures can improve the algorithm’s performance. Additionally, we will discuss multiple criteria optimization problems and imprecise (stochastically disturbed) objective functions.