Generative Adversarial Networks for students' structure prediction. Preliminary research

Abstract

The effectiveness of the university's functioning and its organizational culture can be improved thanks to the use of machine learning. At Universities, the context of student anticipation is very important from the point of view of the fundamental planning and control functions associated with this specific form of management. The purpose of this study is to present the results of an experiment involving the prediction of student structure (attributes of students and their activities) based on the use of a machine learning solution and comparing them against real data obtained from a registry system of a European public institution of higher education in economic sciences. At universities, there is a clear need to support various components of system management. The experiments revealed that -for 11 out of the 48 examined datasets -the Percentage Similarity Index was in excess of 75% but was decidedly lower for the remaining sets (with 18 sets assessed below the margin of 50%).