Prediction of Unbalanced Financial Risk Based on GRA-TOPSIS and SMOTE-CNN

Table 6

Models classification results comparison.

Model name	Macro_Precision	Macro_Recall	Macro_F1	Accuracy
GRA-TOPSIS and SMOTE-CNN	0.9830	0.9830	0.9830	0.9857
GRA-TOPSIS and SMOTE-CNN (without FL)	0.9612	0.9590	0.9599	0.9602
GRA-TOPSIS and CNN	0.9568	0.9263	0.9401	0.9576
Kmeans-CNN	0.9510	0.9437	0.9466	0.9490
Kmeans-SVM	0.9498	0.7669	0.8139	0.9052
Kmeans-KNN	0.3860	0.2943	0.3008	0.3681
Kmeans-decision tree	0.8972	0.7984	0.8262	0.9261
Kmeans-BPNN	0.9729	0.9383	0.9537	0.9717