We have proposed a new SMOTE variant, called SMOTE_AMSR. Although we did not test it on the 104 imbalanced datasets, our algorithm performed very well from the results of those used datasets. Thus, we want to participate in the volunteery competition. SMOTE_AMSR uses the sparsity evaluation function to score each instance of the minority class. Additionally, it uses the geometric topology-based interpolation method to expand the diversity of the minority class
We have proposed a new SMOTE variant, called SMOTE_AMSR. Although we did not test it on the 104 imbalanced datasets, our algorithm performed very well from the results of those used datasets. Thus, we want to participate in the volunteery competition. SMOTE_AMSR uses the sparsity evaluation function to score each instance of the minority class. Additionally, it uses the geometric topology-based interpolation method to expand the diversity of the minority class