Fig. 3
From: An approach for handling imbalanced datasets using borderline shifting
Illustration of the SMOTE-ENN process for identifying and removing noisy or overlapping samples based on their nearest neighbors. After applying SMOTE to generate synthetic minority samples, the Edited Nearest Neighbor (ENN) rule examines each instance and its surrounding neighbors using Euclidean distance. Samples whose class labels differ from the majority of their neighbors are considered noisy or misclassified and are subsequently removed. This combined oversamplingcleaning approach enhances boundary definition and reduces class overlap before model training27.
