Konrad Krawczyk
supervisor: Jarosław Arabas
Evolutionary computation is a nature inspired approach to optimization which is the process of getting the most out of something and making it better. During this process we are looking for the best or optimal solution to the defined problem. For evolutionary algorithms, standing problems are usually complex functions where with iterative evaluations we have to find the best solution. The general idea behind this whole process is that if a biological evolution can produce something as amazing as humans over many generations, we should be able to use similar process artificially to evolve optimal solutions for various defined problems.
Evolutionary algorithms often use functions which evaluations are computationally expensive. In that case there are methods such as surrogate models, that can improve efficiency of this process. Surrogate models are used to substitute “black-box models” which are computationally expensive. In the evolutionary computation we can try to approximate complex functions with surrogate models which results should also provide an optimal solution for the defined problem with computational savings. With that background, the problem could be defined as looking for how to optimize the optimization.