On the Possibility of Developing a System for Predicting Students Academic Performance Using Machine Learning Methods

This article considers the possibility of developing data mining techniques and applying machine learning models to identify hidden dependencies that uniquely characterize different categories of students in order to predict their academic performance. The application of machine learning in this subject area is conditioned by their ability to analyze large amounts of data, obtaining, as a result, generalizing models capable of classifying the objects of study and predicting the required indicators. The development of such a system will allow, based on the analysis of a priori data and current indicators, to predict the expected performance of students in disciplines of different orientation, providing information on the categories of performance in a visual form.