Dynamic Ensemble Machine Learning Classifier Based Credit Card Financial Risk Management and Prediction

The goal is to efficiently manage financial market which evaluate the credit card and reduce the risk of financial enterprises. To address systemic risk arising from insufficient credit scoring in existing financial market. Therefore, this research proposes a Dynamic Ensemble Machine Learning (DEML) classifier for credit card financial risk management and prediction. The ensemble algorithm includes Random Forest (RF), Extreme Gradient Boosting (XGBoost), Support Vector Machine (SVM) and Artificial Neural Network (ANN). These ensemble classifier handles intricate data patterns effectively and capturing connection thereby reducing overfitting issues. It enhances the prediction stability and robustness which leads better risk assessment and reduced false negatives and positives. Moreover, it manages the imbalanced data in which the common financial risk prediction through aggregating prediction among reliable result. The f-measure, precision, accuracy and recall are taken to estimate DLEM classifier performance. The DELM classifier achieves 98.82% f-measure, 98.37% precision, 99.16% accuracy and 98.73% recall which is better than Deep Belief Network (DBN).