How to Combat Imbalanced Classes in Classification Prolbem?
1. More data
2. Metrics
2.1. Confusion Matrix
2.2. Precision
2.3. Recall
2.4. F1 Score
2.5. Kappa (Cohen kappa)
2.6. ROC Curves
3. Re-Sampling
3.1. Over-sampling in the minority label data
3.2. Under-sampling in the majority label data
4. SMOTE(Synthetic Minority Over-sampling Technique) - artificially generate new examples of the minority class using the nearest neighbors of these cases