Addressing Class Imbalance in Loan Default Prediction: A Cluster-Based Synthesis Approach

Keyphrases

• Augmented Dataset 50%
• Binary Classification 50%
• CatBoost Classifier 50%
• Class Imbalance 100%
• Classification Model 50%
• Cluster-based 100%
• Clustering Methods 50%
• Data Synthetization 50%
• Decision Tree 50%
• Default Prediction 100%
• Ensemble Learning 50%
• Family Loans 50%
• Freddie Mac 50%
• Hierarchical Clustering 50%
• Loan Default 50%
• Loan Default Prediction 100%
• Minority Class 50%
• Model Bias 50%
• Model Performance 50%
• Multi-layer Perception 50%
• Oversampling Technique 100%
• Performance Evaluation Index 50%
• Performance Prediction 50%
• Preprocessing Techniques 50%
• Public Loan 50%
• Simpson's Paradox 50%
• Single-family 50%
• State-of-the-art Techniques 50%
• Synthetic Approaches 100%
• Synthetic Minority Oversampling Technique (SMOTE) 100%
• Testing Sample 50%
• Training Samples 50%
• XGBoost Classifier 50%

Computer Science

• Binary Classification 50%
• Class Imbalance 100%
• Classification Models 50%
• Classification Problem 50%
• Decision Trees 50%
• Ensemble Learning 50%
• Evaluation Metric 50%
• Experimental Result 50%
• Extreme Gradient Boosting 50%
• Generalizability 50%
• Hierarchical Clustering 50%
• Minority Class 50%
• Performance Evaluation 50%
• Performance Model 50%
• Predictive Performance 50%
• Preprocessing Technique 50%