Maestría en Analítica Aplicada · 2026

Clasificación y predicción del archivo de procesos judiciales en Colombia mediante técnicas de machine learning

Resumen

The Colombian judicial system is facing a crisis of efficiency due to case overcrowding, lengthy trials and under-utilization of procedural data. This research developed a data architecture based on the medallion approach, which allowed to integrate and analyze judicial records from open data. From the exploratory analysis, patterns associated with the procedural stage, sectional stage, type of crime and other variables that affect the process archive were identified. With this information, classification models were trained using Machine Learning techniques to predict the probability of filing a court case.