An Educational Data Mining Model based on Auto Machine Learning and Interpretable Machine Learning

This paper proposes a new Data Mining Educational Model for knowledge extraction with a minimum presence of data scientists, a scarce and determinant human resource for the application of machine learning in Data Mining. The model allows generating data sets semi-automatically, obtaining an optimized algorithm through automatic machine learning (AutoML) and explaining the results with interpretable machine learning (IML). It is applied in the field of University Education and implements a process with three general stages (Data Analysis and Integration, Data Modeling and Results Evaluation) and is validated by means of a friendly prototype to non-expert users with data obtained from an Argentine Public University. With this proposal we aim to allow universities to draw conclusions on complex problems, requiring a minimum number of data science experts and providing a framework for both end users and legal entities to be informed of the results generated.