Machine Learning for Evaluation and Prediction of Financial Profiles in the Subsidized Regime in Colombia

Tomás FONTALVO-HERRERA, Enrique DELAHOZ-DOMINGUEZ and Orianna FONTALVO-ECHAVEZ

Abstract:

This research evaluates, classifies and forecasts financial profiles in the subsidized regime in Colombia. The above is supported by a theoretical framework related to multivariate calculation and machine learning, specifically Random Forest and GLMNET. The work is of qualitative, evaluative and predictive approach. After the application of the different tools, it was found that two business financial clusters were discovered, which could be predicted through the Random Forest and GLMNET algorithms with 100% accuracy

nsdlogo2016