Dr Mateusz Krzysztoń

Research interests: cybersecurity, machine learning, security of artificial intelligence

Mateusz Krzysztoń heads a team that conducts research and development work on the application of machine learning in cybersecurity. The first area of research concerns the detection of unknown threats in cyberspace, including, among others, new malicious applications on mobile devices or new types of attacks on ICT systems by observing network traffic. The second is to study and evaluate the security of machine learning models, particularly the vulnerability of these models to attacks.

The team’s biggest success is the development of a lightweight detector for new malicious applications on Android. The detector runs entirely on the client’s mobile device, so that information about the apps the user installs does not go beyond his device. The solution is currently being implemented as part of the BotSense system widely used by banking institutions in Poland.

Selected Publications

Articles

Giuseppe Stragapede et al., "IEEE BigData 2023 Keystroke Verification Challenge (KVC)", IEEE International Conference on Big Data (BigData), Sorrento, Italy, 2023, 6092-6100.
Mateusz Krzysztoń, Bartosz Bok, Marcin Lew, Andrzej Sikora, "Lightweight On-Device Detection of Android Malware Based on the Koodous Platform and Machine Learning", Sensors, 22(17), 2022, 6562.
Joanna Kołodziej, Mateusz Krzysztoń, Paweł Szynkiewicz, "Anomaly Detection in TCP/IP Networks", Communications of the ECMS, ECMS 2023, 37th Proceedings, Volume 37, Issue 1, 2023,

Book Chapters

Mateusz Krzysztoń, "Weryfikacja wiarygodności systemów w erze uczenia maszynowego", Cyberbezpieczeństwo AI. AI w cyberbezpieczeństwie, Warszawa: NASK PIB, 2023, 45–58.
Mateusz Krzysztoń, Marcin Lew, Michał Marks, "NAD: Machine Learning Based Component for Unknown Attack Detection in Network Traffic", Cybersecurity of Digital Service Chains. Challenges, Methodologies, and Tools, Switzerland: Springer Cham, 2022, 83–102.