From November 7 to 18, 2023, the project shift "Trusted Artificial Intelligence" will take place in the scientific center «Information Technologies and Artificial Intelligence». The shift will include projects from a group of laboratory employees: Alexander Beznosikov, Dmitry Kovalev, Alexander Lobanov, Darina Dvinsky and Alexander Gasnikov.
Students of the senior courses of bachelor's and specialist's degree, master's degree and postgraduate training programs are invited to participate, studying in the following groups of specialties: mathematics, applied mathematics and computer science. You can apply until October 1, 2023.
The program includes:
- methods for detecting and countering machine learning vulnerabilities, including adversarial attacks, attacks with the introduction of bookmarks and malicious code, etc.;
- methods for identifying and combating model obsolescence due to changes in distributions in input data (concept drift, feature drift);
theory and practice of creating models resistant to attacks; - approaches to interpretability of machine learning models;
- creating trusted frameworks for training neural networks;
- static and dynamic analysis of components of intelligent systems.
Shift participants will also be able to get acquainted with the developments of the Research Center for Trusted Artificial Intelligence of the ISP RAS, as well as participate in the creation of a cloud platform for the development of trusted intelligent systems. There will also be lectures/projects from famous people who lead their groups, for example, from Ivan Oseledets, Dmitry Vatolin and others. The main organization will be the Center for Trusted AI of the ISP RAS.
Learn more about the selection stages, requirements for participants and conditions on the Sirius website.
Photo source: Sirius website.