IMPROVING MODEL PERFORMANCE FOR PREDICTING EXFILTRATION ATTACKS THROUGH RESAMPLING STRATEGIES

Keyphrases

• Class Distribution 50%
• Class Imbalance 100%
• Class Separability 50%
• Comprehensive Framework 50%
• Cybersecurity 50%
• Cybersecurity Applications 50%
• Exfiltration 100%
• F1 Score 50%
• Geometric Mean 50%
• Imbalanced Data 50%
• Improving Model 100%
• Labeled Data 50%
• Logistic Regression 50%
• Majority Class 50%
• Metric Accuracy 50%
• Minority Class 50%
• Model Performance 100%
• Nave Bayes 50%
• Performance Enhancement 50%
• Performance Metrics 50%
• Poor Generalization 50%
• Precision-recall 50%
• Random Forest 50%
• Resampling 100%
• Separability 50%
• Skewed Datasets 50%
• SMOTE-ENN 100%
• SMOTETomek 50%
• Stability of Distributions 50%
• Support Vector Machine 50%
• Synthetic Minority Oversampling Technique (SMOTE) 100%
• Variant Impact 50%
• XGBoost 50%

Computer Science

• Class Distribution 50%
• Class Imbalance 100%
• Cybersecurity 100%
• Enhance Performance 50%
• Extreme Gradient Boosting 50%
• Imbalanced Data 50%
• Logistic Regression 50%
• Majority Class 50%
• Minority Class 50%
• Nave Bayes 50%
• Performance Metric 50%
• Performance Model 100%
• Random Decision Forest 50%
• Support Vector Machine 50%