The article is focused on the problem of early prediction of students’ learning failures with the purpose of their possible prevention by timely introducing supportive measures. We propose an approach to designing a predictive model for an academic course or module taught in a blended learning format. We introduce certain requirements to predictive models concerning their applicability to the educational process such as interpretability, actionability, and adaptability to a course design. We test three types of classifiers meeting these requirements and choose the one that provides best performance starting from the early stages of the semester, and therefore provides various opportunities to timely support at-risk students. Our empirical studies confirm that the proposed approach is promising for the development of an early warning system in a higher education institution. Such systems can positively influence student retention rates and enhance learning and teaching experience for a long term.

Problem statement. Learning analytics is an emerging scientific field, which studies learners and learning process based on data from digital environment. The aim of the study - to observe the development of learning analytics, its prospects and limitations and detecting the state of art of this scientific field in Russia. Methodology . The study is based on context analysis of scientific articles on the topic in the public domain. Special attention is given to reviewing scientific publications of Russian-speaking authors devoted to analytics of education data and the implementation of learning analytics tools in the educational process. Results . The research detects the global directions of learning analytics development and its problematic aspects. It provides the quantitative and qualitative analysis of scientific publications of Russian-speaking authors and identifiers the most popular research questions in the learning analytics field. It proposes the author’s vision of the hierarchy of directions for learning analytics development, consisting of the research aspect, the environment transformation aspect and the legal regulation aspect. The national initiatives in the digitalization of education are briefly discussed. Conclusion . A certain lag in the level of development of learning analytics in Russia from the global one is revealed. At the same time, there is a noticeable increase in interest to this area among individual researchers, educational institutions and at the state level, which allows us to count on positive changes.

Student retention prediction is one of the most important problems of learning analytics. In the global scope research on the topic for higher education is rather extensive, there are cases of successful implementation of education support services in universities. The literature analysis shows of the growing interest in this problem in the Russian scientific and pedagogical community. At the same time, the specifics of Russian education does not allow direct transfer of foreign experience into the domestic educational system.The study reveals that a significant contribution to predicting student retention can be made by models for predicting academic performance in educational courses of the curriculum. The authors propose a structural model of a system for predicting academic performance, which includes a universal model based on generalized indicators of the digital footprint, a course-based model that takes into account the specifics of learning in a particular discipline, and a model based on the student’s educational profile.In the empirical study we trained 5 models for early prediction of interim assessment grades based on the universal indicators of the LMS Moodle student digital footprint. The most accurate model, especially in the first half of the semester, turned out to be ensemble-averaging models of logistic regression, random forest and gradient boosting. It was found that universal models are effective for detection of at-risk students in the discipline, the directions for further improvement of the universal model of performance prediction were determined and conditions for scaling the proposed approach to create a prognostic system for student retention to other educational institutions were formulated.

The article is dedicated to the problem of effective presentation of theoretical material by e-courses in mathematical disciplines, which arises due to the peculiarities of teaching mathematics, and of such hypophysical processes as attention and memory. We consider the pedagogical principles of microlearning and combining theory with practice, following which allows to increase the degree of theoretical material digestion by students. We propose implementing these principles with the help of training lectures, in which portions of theoretical material alternate with test tasks. We describe an experiment for studying the effectiveness of training lectures in teaching students using the electronic course “Probability theory” created in the Learning Management System Moodle. We examine the effectiveness of training lectures in the educational process in comparison with standard electronic lecture, which is usually presented as a multipage text file.