Classification of university teacher performance using machine learning

This work aimed to determine and propose a teacher classification methodology based on criteria evaluated by students using a methodology based on Machine Learning. For this purpose, 114 classes of the Faculty of Engineering underwent evaluation, where Academic Satisfaction (SA) and Net Promoter Score (NPS) surveys were administered to 3,532 students. In the unsupervised analysis, 4 clusters were determined based on the k-means algorithm with an R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> of 0.88, which showed relationships between the criteria evaluated by the students. Finally, using a supervised algorithm, such as the K-Nearest Neighbors Classification, the model was adjusted to 3 scales, with which the proposed teacher classification was constructed. These results allow us to propose 3 teacher scales focusing on a continuous improvement process.