Predictive Analytics in Digital Education: Analysis and Evaluation of Students’ Learning Success

Abstract

Problem and Aim. The article actualizes the problem of modeling the learning process within electronic educational environment, based on the use of educational analytics data. The purpose of the article is to reveal and substantiate the possibilities of using predictive analytics methods for analyzing and assessing the of students’ learning success based on their current educational results. Methodology. Theoretical, empirical and mathematical methods were used during the research. The construction of the predictive model was carried out on the basis of the technologies of intellectual analysis and processing of big data (Data Mining). Research results. The relevance of the use of methods for analyzing big data in the educational sphere for improving the quality and efficiency of the educational and pedagogical process has been substantiated. The definitions of the technology of data mining are given, its types are considered (descriptive, diagnostic, predictive and prescriptive). The machine learning methods used in predictive analytics algorithms are described. Examples of the using predictive analytics methods in various fields of activity are given. An algorithm for designing and developing a predictive model is described. On its basis, a predictive model is proposed for analyzing and assessing the success of student learning, implemented as part of an experimental study within training module “Optimization Methods”. To assess the quality of the model, the methods of linear and logistic regression, deep neural networks and a decision tree were applied. In conclusion, it is concluded that the development and implementation of a model based on a set of methods gives a synergistic effect – the ability to maximally take into account the peculiarities of a student’s work with educational material and the degree of its assimilation. And as a result, it allows you to increase the level of training of students, improve the quality of the educational material provided in an electronic environment, build an individual trajectory of learning and help eliminate the problem of student expulsion through timely action.