A Novel SMOTE-Based Classification Approach to Online Data Imbalance Problem

Figure 1

The distribution of the offline chessboard-shaped data (a) before and (b) after the first granulation and (c) the result after using SMOTE.

(a)

(b)

(c)