Experiments on Imbalanced Data
Preprocessing methods
Oversampling methods
- Random oversampling
- SMOTE
- Borderline-1 SMOTE
- Borderline-2 SMOTE
- Borderline SMOTE-SVM
- SVM-SMOTE
Undersampling methods
- NearMiss-1
- NearMiss-2
- NearMiss-3
- Edited Nearest Neighbor
- Repeated Edited Nearest Neighbor
- Tomek LinksEasy Ensemble Classifier
- Condensed Nearest Neighbor
Imbalanced data classification algorithms
- Logistic Regression
- Random Forest
- Easy Ensemble
- Super Vector Machines
References
The dataset used: Credit Card Fraud Detection.