Decision Tree Learning for Credit Risk Assessment in Banking

Credit risk assessment is one of the most important banking functions. It provides a basis for estimating the default risk of borrowers, which thereby influences lending decisions in the most appropriate manner. The present work reviews a Credit Assessment based on Decision Tree, beginning with data preprocessing, model training, hyperparameter tuning, and model performance benchmarking. The paper studies the possibilities, efficiencies, and limitations of the use of Decision Trees with respect to credit risk while making comparisons with more complex ensemble methodologies. Considerations regarding practical deployment are made, focusing on possibly realistic scenarios regarding online credit scoring systems. Performance evaluation shows the Decision Tree approach achieved 85% classification accuracy