Enhanced Bankruptcy Prediction Using Hybrid Machine Learning Techniques with Decision Tree Classifier on the Polish Dataset

Abstract

Bankruptcy forecasting is vital for monetary stability and threat management. This look at improves the accuracy of financial disaster prediction using Decision Tree techniques on an imbalanced Polish dataset. The dataset affords demanding situations normal in actual-world economic facts, along with magnificence imbalance and noisy capabilities. Our method applies Decision Tree, a easy yet effective system studying model, to address these demanding situations successfully. Experimental outcomes show extensive improvements in precision, bear in mind, and F1-rating metrics when as compared to conventional techniques. Insights received from feature importance analysis offer a deeper know-how of monetary signs that pressure financial disaster. This research contributes to the field by way of offering a sturdy framework the usage of Decision Tree that is adaptable to imbalanced datasets, offering realistic insights for monetary institutions to mitigate financial disaster risks efficaciously.