Using predictive methods for assessing academic performance of graduate students based on LMS platform data

The paper assesses the prospects for using LMS platform data and modern methods of statistical data analysis - decision trees - for a proactive assessment of the academic performance of Russian graduate students and predicting the likelihood of them successfully defending their thesis. The paper postulates the problem of only a small proportion of Russian graduate students successfully complete postgraduate studies and defend their thesis. In the process of finding a solution to this issue, the author conducts a review of relevant studies to identify student performance factors, provides a brief comparison of the available statistical tools and data sources, examines a case of practical application of the selected tools by western researchers. The author comes to the conclusion that a practical study is necessary to access the possibility of using LMS data along with decision trees to predict Russian graduate students’ academic performance.