Quantum machine learning and security on the quantum internet - Nana Liu

The success of the modern internet relies in no small part on understanding the interplay between computation and security. More recently this area has also seen contributions from machine learning, including spam filters and malware detection. With the rising prevalence of machine learning algorithms, it is also important to address whether new security issues arise. At the same time, as quantum technologies become more accessible, these same security concerns for quantum data is expected to become more important in a future quantum internet. It then becomes intriguing to study the overlap between security and machine learning when enhanced by quantum resources.
Machine learning applications often require training or test data that originate from remote data centres or sensors. This decentralised set-up, which forms one of the application of the quantum internet, opens the door to adversaries that could exploit existing vulnerabilities in those algorithms. This incites us to ask some basic questions: What is the vulnerability of quantum machine learning algorithms? How can we exploit quantum machine learning for security purposes? We take some first key steps in addressing these questions and look towards the future of this intersection between computation and security on quantum data.


Join Zoom Meeting https://ucalgary.zoom.us/j/95197013180 Meeting ID: 951 9701 3180