Research groups
Department of Computer Science
Unconventional Algorithms and Computation
The group's research interests focus heavily on cybersecurity, artificial intelligence, and bio-inspired computing methods, with a particular emphasis on the application of AI techniques to cyber defence and offence. One significant area they explore is the use of swarm intelligence and evolutionary algorithms in the development of novel cybersecurity systems, such as the "Swarm Virus" concept, which merges ideas from virus behaviour in nature with digital virus spread to propose next-generation paradigms for malware and antivirus design.
Group work also delves into the intersection of AI and cybersecurity, where they study AI-driven intrusion detection systems (IDS), utilising machine learning and deep learning models to enhance the detection of sophisticated cyber threats. These include employing AI for anomaly detection in networks, enhancing the efficiency of IDS, and improving the classification of malware through advanced algorithms such as genetic algorithms and neural networks.
Additionally, our recent publications investigate the potential of AI in automating cyber-attacks, exploring both defensive and offensive applications. By utilising AI, they aim to create more resilient security systems while also acknowledging the increasing risks posed by AI in the hands of attackers. Their work is central to understanding how to balance AI's dual role in enhancing cybersecurity while also preventing its misuse in cyber warfare.
We also developed a novel approach to malware visualisation, utilising dynamical analysis, complex networks, and fractal geometry such as Julia sets. This method converts malware samples into visual representations, classifying them as malware or goodware using deep learning networks. Based on an extensive database of over 11 million samples provided by ESET, the experiments show potential for improved analysis and classification. This innovative approach to malware visualisation through fractal geometry opens up new possibilities for advanced malware detection and categorisation.
Over the last two years, we have also been actively conducting research in quantum computing. The Navy explores the use of swarm intelligence algorithms, such as iSOMA, to design optimised quantum circuits. Experiments have shown that iSOMA can creatively generate efficient quantum circuits while minimising resource usage, demonstrating potential in complex quantum circuit design using the IBM Qiskit environment.
Research topics:
- Unconventional computational methods (bioinspired computation, fractal geometry, deterministic chaos, cellular automata,...) and their applications in control, modelling, optimisation and identification
- Computer and communication security
- AI in cybersecurity
- AI and quantum algorithms
- AI methods in Esports
Group leader: prof. Ing. Ivan Zelinka, Ph.D.
Laboratories
The NAVY group focuses on research of unconventional algorithms and computational approaches in cybersecurity, artificial intelligence, and biologically inspired methods. A key direction is the use of swarm intelligence and evolutionary algorithms in developing security systems, such as the “swarm virus” concept, which links biological and digital behaviour. Another central area is intrusion detection (IDS) based on machine learning and deep learning, anomaly detection, IDS efficiency improvement, and malware classification using genetic algorithms and neural networks. We also explore large language models (LLMs, e.g., GPT) for defence, threat identification, and penetration testing. Our publications address the automation of cyberattacks using AI in both defence and offence, while stressing the need to balance AI’s dual role. We developed a novel malware visualisation approach combining dynamic analysis, complex networks, and fractal geometry, validated on a dataset of over 11 million samples from ESET. This opens new possibilities in detection and classification. In recent years, we have also focused on quantum algorithms, particularly the use of swarm methods (e.g., iSOMA) for designing optimised quantum circuits, demonstrating their ability to generate efficient structures with minimal resources.
Our research group boasts a state-of-the-art PC lab dedicated to AI and cybersecurity research, where we conduct advanced simulations, experiments, and model training in a secure and cutting-edge environment. Beyond this, we are also actively involved in the Quantum Computing Laboratory at (https://www.it4i.cz/en/research/research-laboratories/quantum-computing-lab) IT4Innovations, a leading supercomputing centre. This collaboration allows us to tap into quantum resources, pushing the boundaries of quantum computing and its applications in cybersecurity and artificial intelligence. Together, these labs form the backbone of our innovative research.
Gallery of laboratories
