Feature selection and its combination with data over-sampling for multi-class imbalanced datasets

Keyphrases

• Feature Selection 100%
• Class Imbalance 100%
• Ensemble Feature Selection 100%
• Over Sampling 100%
• Class-imbalanced Datasets 100%
• Feature Selection Methods 83%
• Combined Algorithm 50%
• Synthetic Minority Oversampling Technique (SMOTE) 50%
• Classification Performance 33%
• Single Feature 33%
• Sampling Algorithm 16%
• Multiple Feature Selection 16%
• Learning Model 16%
• Result-oriented 16%
• Classification Accuracy 16%
• Feature Dimension 16%
• Effective Learning 16%
• Synthetic Instances 16%
• XGBoost 16%
• Best Combination 16%
• Filtering Algorithm 16%
• Outperformance 16%
• Minority Class 16%
• Rebalance 16%
• Embedding Algorithm 16%
• Performance Difference 16%
• Filter-Wrapper 16%
• Extreme Gradient Boosting(XGBoost) 16%
• Sampling Step 16%
• Wrapper Algorithm 16%

Computer Science

• Feature Selection 100%
• Feature Extraction 100%
• Extreme Gradient Boosting 23%
• Classification Performance 15%
• Research Objective 15%
• Experimental Result 7%
• Classification Accuracy 7%
• Feature Dimension 7%
• Single Feature 7%
• Minority Class 7%
• Performance Difference 7%
• Sampling Algorithm 7%