Software for university students based on machine learning

The import substitution issue in IT sector has become highly relevant nowadays, leading universities which offer educational programs in this field to search for the usual products alternatives. One possible solution to the problem is self-development. In this paper, we aimed to create software that can be used for laboratory work in disciplines related to the study of machine learning algorithms. Authors utilized methods of stochastic gradient descent, high-level programming, and comparative analysis to develop the software. It was tested, and positive results have been obtained, confirming the possibility of this solution practical usage.