Forecasting the risk of expulsion from the university as a result of academic failure of the student

The issues of overcoming academic failure and, as its result, the risk that the student will not finish his education are relevant. In this article, the author proceeds from the assumption that academic success, especially in higher education, is influenced by several groups of factors, the identification and timely monitoring of which will be the basis for the development or correction of an individual learning route that contributes to academic success and prevents the risk of student expulsion. In the process of theoretical and empirical longitudinal research, which lasted forsix years, four groups of academic success factors were identified in a sample of 579 medical university students (from the moment of admission to the university to its graduation): — external unmodifiable (regulated conditions of education), -external modifiable (organization of education at the university), — internal unmodifiable internal modifiable. A discriminant function was constructed for deduction after the first year of medical school, in which the indicators of maladaptation influenced the deduction as a key factor. The sensitivity of the formula turned out to be good, more than 75 %. Adding to the decision-making algorithm information about the grades obtained from the results of training for the first semester, allowed using the construction of a decision tree to improve the overall accuracy of the forecast to 94,6 %. As the Decision Tree shows, the main indicator of deduction is the average grade for the completed semesters, then, if it is high, then the average score for the completed semesters and the total score for admission. Monitoring of the proposed factors of academic success makes it possible to predict the risk of failure among students and provide them with timely assistance so that they continue their studies.