THE EMPLOYMENT OF MACHINE LEARNING ALGORITHMS FOR PREDICTION IN LEARNING ANALYTICS AND EDUCATIONAL DATA MINING WITHIN THE CONTEXT OF HIGHER EDUCATION

This paper presents a review of the literature from the last five years on predictive methods of Learning Analytics and Educational Data Mining based on Machine Learning algorithms. The primary selection criterion for the papers analyzed was to identify those that use Machine Learning algorithms to predict outcomes in the areas of Learning Analytics and Educational Data Mining in the context of higher education. It is important to highlight that there are no universal guidelines or protocols for predicting outcomes in education, including higher education. The methodology used for such predictions depends primarily on the target variable and the type of input data used. Twenty-five papers from the Web of Science CC and Scopus citation databases were included in the detailed analysis. Six research questions were used to examine what is being predicted in higher education, what input data were used, how many Machine Learning algorithms were used in the research, and which were most effective. In addition, the research looked at what other predictive modeling techniques were mentioned and whether the programming environment used for prediction was mentioned.