Interpretable Electronic Transfer Fraud Detection with Expert Feature Constructions

Keyphrases

• Detection Performance 100%
• Fraud 100%
• Fraud Detection 100%
• Modus Operandi 100%
• Feature Construction 100%
• Electronic Transfer 100%
• Decision Tree 50%
• Tree-based 50%
• Transaction Data 50%
• Vulnerability 50%
• System Process 50%
• Result Prediction 50%
• Feature-based 50%
• Ranking Method 50%
• Rule-based Approach 50%
• Recency 50%
• Temporal Features 50%
• Kolmogorov-Smirnov Test 50%
• Machine Learning Techniques 50%
• Financial Expertise 50%
• Recall Rate 50%
• Causal Effect 50%
• Precision Rate 50%
• Financial Institutions 50%
• Time-varying Data 50%
• Clearance Rate 50%
• Banking 50%
• Traditional Rule 50%
• Labor Costs 50%
• Financial System 50%
• XGBoost 50%
• Modeling Time 50%
• Segmentation Feature 50%
• Financial Fraud 50%
• Extreme Gradient Boosting 50%
• Frequency Features 50%
• Light Gradient Boosting Machine 50%
• Financial Characteristics 50%
• Fraudsters 50%
• Low Clearance 50%
• Financial Organizations 50%

Computer Science

• Detection Performance 100%
• Fraud Detection 100%
• Extreme Gradient Boosting 100%
• Feature Construction 100%
• Electronic Transfer 100%
• Temporal Feature 50%
• Transaction Data 50%
• Interpretability 50%
• Frequency Feature 50%
• Gradient Boosting 50%
• Machine Learning 50%
• Learning System 50%
• Decision Tree 50%