Porovnání běhů algoritmu diferenciální evoluce pomocí sítí

Abstract

This bachelor thesis deals with the recording the runs of the algorithm differential evolution. Next, runs are converted to the networks and a comparison is made between the better and the worst ones. For this purpose, two tools have been implemented in this thesis. Principles of biologically inspired computations and evolutionary algorithms with a detailed focus on algorithm differential evolution are described. The following chapter defines the concepts of oriented and weighted networks that are needed for subsequent analysis. Before the data evaluation, the tools used here are described. In the last chapter, the thesis focuses on the analysis of generated data and the search for differences between better and worse runs of differential evolution algorithm.