Artificial Intelligence:

This research seeks to apply the use of machine learning algorithms (Artificial Intelligence) to the prediction of different variables related to the phenomenon of corruption in the municipal administrations of Colombia. The ultimate goal is to develop risk maps that serve as early warnings to control agencies to identify and prevent acts of corruption in advance. The algorithms that are trained and evaluated are Random Forest, Gradient Boosting with a linear regression base model and Bradient Boosting with a decision tree base model. The final predictions are grouped into clusters that allow identifying which municipalities present a higher risk of corruption in similar municipalities.