Machine Learning for Supporting Cybersecurity

Automatic and comprehensive methods are crucial to protect the complex systems connected to the Internet. Our research focuses on exploring network measurements and machine learning to assist on the cybersecurity. We develop and apply methods to collect the right information from suitable sources in the network to support cybersecurity tasks. We leverage machine learning approaches, such as unsupervised clustering methods, anomaly detection algorithms and deep neural networks, to build up systems able to identify attacks, track the presence and evolution of malware and viruses, and highlight anomalous behaviors that may be associated with abuses, misuse and misconfigurations of the network.


ERC Sector:

  • PE7_8 Networks (communication networks, networks of sensors, robots...)


  • Network security
  • Machine learning
  • Internet ceasurements
  • Cybersecurity

Research groups