We previously reported that ISU ranked among the top 10 Russian universities by the number of winners in the Vladimir Potanin Foundation’s competition for professors of Master’s degree programmes. Four ISU representatives became winners.

Today we will talk about the project “Workshop on methods of analysis and deep machine learning for master’s degree students without specialized mathematical education” by Oleg Bernhardt, Candidate of Physical and Mathematical Sciences, Associate Professor of the Department of Natural Sciences at the Faculty of Business Communications and Computer Science.

The project focuses on redesigning the machine learning course for graduate students without specialized mathematical background. In recent years, there has been a technological breakthrough in large language models (ChatGPT, o1, DeepSeek, and others). Redesigning the existing course is necessary to cover new areas, it also includes adding units related to generative and multimodal models, as well as the mathematical and algorithmic methods necessary for understanding and development of the mentioned models.

Oleg Bernhardt:

The difficulty in implementing such a project lies in creating a set of small, illustrative tasks, in solving which students will be able to create and train their own neural networks with modern architectures within the time limit of one class. On the other hand, the complexity of this project lies in explaining the mathematical knowledge in the amount necessary for the successful solution of such problems. As a result of the project, a basic set of tasks, examples and datasets for the reorganized course will be created, suitable for use and distribution in teaching master’s degree students without specialized mathematical education. Methodological materials to go with the course in text and short video form will also be created and made publicly available.

A part of the simple tests in the course will be implemented as test assignments of an online course on the Forlabs online learning system used at the Faculty of Business Communications and Computer Science.

It is expected that the updating and expansion of the course will improve the quality of training of the graduate students, broaden their horizons and expand their experience in the field of current machine learning methods. This will increase the competitiveness of ISU educational programmes in the field of training graduates with skills in machine learning and give a competitive advantage to the graduates.