SMOTE: Synthetic Data Augmentation for Tabular Data

An exploration of SMOTE and some variants like Borderline-SMOTE and ADASYN

Figure 1. SMOTE, Borderline-SMOTE and ADASYN representation | Image by author | Icons taken from freepick

The class imbalance problem occurs when there is no balanced distribution among classes. The intuition to solve such a problem is to add more data to the minority class to generate a balance among the classes however, in real machine learning systems, it…