Přehled a implementace vybraných shlukovacích algoritmů založených na metodě "random walk"

Abstract

The aim of this diploma thesis describe, implement and summarize few selected clustering algorithms, which are based on the random walk method. On the beginning of the thesis we focus on the theoretical basics, which will be needed to know for understanding rest of this thesis. Furthermore, in theory, we explain and analyze the selected algorithms and perform experiments over different types of smaller data collections. The next section will focus on the implementation part, which will explain the procedures and the resulting solution implementations described clustering algorithms. At the end of this thesis we properly perform and evaluate experiments over real-world data.