Model for discovering knowledge about academic and administrative aspects for students at driving schools in San Juan De Pasto

This paper proposes a comprehensive methodology for knowledge discovery in databases (KDD) applied to driving schools. The usefulness of clustering algorithms such as K-means and K-prototype to identify patterns in administrative and academic procedures was explored. During the study, three main stages were developed: process characterization, experimental design based on machine learning, and evaluation of the generated models. The results showed that K-prototype is particularly effective in handling mixed data, providing key recommendations to optimize both training processes and internal management. In addition, an application was designed to implement the model, highlighting the impact of educational data mining on dynamic analysis and informed decision making.