The seminar was organized by the Laboratory of Mathematical Methods and Artificial Intelligence, created within the framework of the Center for Training Researchers and Developers in the field of artificial intelligence (director of  the Center M.I. Karyakin).

The seminar program included two key reports.

D.B. Rokhlin, Ph.D., made a presentation on "Learning with cores: a general introduction." The report presented the basic concepts of statistical learning theory and online optimization theory with an emphasis on nuclear methods.

O.E. Kudryavtsev, PhD, presented a report "On increasing the interpretability of artificial neural networks in computational financial mathematics problems." Probabilistic analogues of universal approximation theorems were considered and a connection was established between continuous infinitely divisible random variables and monotonic neural networks with one hidden layer. Based on these results, a new approach to the design of Monte Carlo methods was developed in combination with artificial neural networks for evaluating options in Levy models.

The seminar was attended not only by researchers and teachers, but also by students and undergraduates of the I.I. Vorovich Institute of Mathematics, Mechanics and Computer Science. The seminar was held in the format of an active discussion, where the participants discussed the prospects of applying the presented methods in various fields of science and practice. The event confirmed the high scientific potential of the Southern Federal University in the field of development and application of modern artificial intelligence methods.