Towards hybrid over- and under-sampling combination methods for class imbalanced datasets: an experimental study

Keyphrases

• Combined Method 100%
• Undersampling 100%
• Class-imbalanced Datasets 100%
• Over-under 100%
• Over Sampling 57%
• Oversampling Technique 42%
• Instance Selection 28%
• Sampling Algorithm 14%
• Performance Improvement 14%
• Machine Learning Techniques 14%
• Selection Algorithm 14%
• Class Imbalance 14%
• Resampling Methods 14%
• Hybrid Mixtures 14%
• Random Forest Classifier 14%
• Effective Model 14%
• Majority Class 14%
• Minority Class 14%
• Synthetic Minority Oversampling Technique (SMOTE) 14%
• Combination Order 14%
• Data Resampling 14%
• Skewed Class Distribution 14%
• Sampling Step 14%
• Imbalanced Domains 14%
• Conditional Generative Adversarial Network (cGAN) 14%

Computer Science

• Combination Method 100%
• Instance Selection 100%
• Sampling Algorithm 100%
• Machine Learning Technique 50%
• Performance Improvement 50%
• Research Objective 50%
• Majority Class 50%
• Minority Class 50%
• Sampling Technique 50%
• Class Distribution 50%
• Random Forest Classifier 50%