The management of the security in the cyberspace requires the use of a variety of tools for the early detection of vulnerabilities, signs of misuse, and intrusions. Machine learning approaches are widely adopted to address cybersecurity tasks for their capability of clustering and classifying events described by a large set of diverse characteristics.
This lecture will start by exploring the meaning of the term security in the cyberspace, through a journey among the various layers such as hardware, operating systems, networks, protocols and application software. Then, different case studies will be presented to show the process that has been followed to use machine learning models for detecting malicious behaviours in the cyberspace. Finally, the approaches to make machine learning techniques resilient against adversaries that aim to disrupt or mislead the detection process will be discussed.