Money laundering through cryptocurrencies and the possibility of their detection

The diploma thesis deals with the possibilities of detecting suspicious transactions that resemble money laundering using cryptocurrencies. The aim of the thesis is to analyse the scenario of money laundering through cryptocurrencies and to propose principles for the detection of these operations. The thesis is divided into several chapters, each of which deals with a specific aspect of the problem - from the introduction to the problem, through the analysis of the current regulations, to the design and implementation of detection methods, including the creation of a prototype application for the detection of suspicious transactions. The main contribution of the thesis lies in the demonstration of the implementation of the proposed suspicious transaction detection methods. The final model achieves high efficiency in identifying suspicious transactions and uses a combination of convolutional graph network, isolation forest, one-class SVM and semi-supervised pseudo-label learning. The contribution of this work is particularly significant for entities receiving digital assets, as it provides insight into how specific tools can be implemented to detect suspicious behaviour through the transfer of cryptocurrencies.