Fuzzy-based Active Learning for Predicting Student Academic Performance
Maria Tsiakmaki, Georgios Kostopoulos, Sotiris B. Kotsiantis, O. Ragos · Proceedings of the 6th International Conference on Engineering & MIS 2020 · 2020
Predicting students' learning outcomes is one of the main topics of interest in the area of Educational Data Mining. To this end, a plethora of machine learning methods have been successfully applied for solving a variety of predictive problems. However, it is of utmost importance to develop accurate and explainable learning models at low cost. Fuzzy logic constitutes the appropriate approach for building models of high performance and explainability. In addition, active learning reduces both the time and cost of labeling effort exploiting a small pool of labeled examples together with a large pool of unlabeled ones in the most efficient way, assuming that an expert provides the true labels of the most informative unlabeled examples during the training process. In this context, the present study introduces a fuzzy-based active learning method for predicting student academic performance. Initially, we provide a comparative study on the efficacy of fuzzy learning in five compulsory courses. Therefore, we evaluate the predictive performance of six classes of fuzzy classifiers on datasets regarding students' online activity in each course. In addition, we propose a fuzzy-based method exploiting the potential of the active learning approach. The experimental results demonstrate the efficiency of the proposed method for the accurate prediction of students at risk of failure.